Pure Drink or Pure Hype?
Principal Findings and Recommendations
Americans increasingly are turning to bottled water, making it a $4 billion-a-year business in the United States.  Millions of us are willing to pay 240 to over 10,000 times more per gallon for bottled water than we do for tap water -- though we probably rarely think of it that way.  However, some bottled water contains bacterial contaminants, and several brands of bottled water contain synthetic organic chemicals (such as industrial solvents, chemicals from plastic, or trihalomethanes -- the by-products of the chemical reaction between chlorine and organic matter in water) or inorganic contaminants (such as arsenic, a known carcinogen) in at least some bottles (see Chapter 3 and our accompanying Technical Report [print report only]).[1a] Moreover, as Chapter 4 documents, bottled water regulations have gaping holes, and both state and federal bottled water regulatory programs are severely underfunded. In Chapter 5 we present evidence that there is substantially misleading marketing of some bottled water, and in Chapter 6 we argue that consumers should be informed about the contaminants found in the water they purchase. NRDC's major findings and recommendations are summarized below.
1. Most bottled water apparently is of good quality, but some contains contamination; it should not automatically be assumed to be purer or safer than most tap water.
Based on available data and our testing, most bottled water is of good quality, and contamination posing immediate risks to healthy people is rare (see Chapter 3 and the Technical Report [print report only]). However, blanket reassurances from the bottled water industry that bottled water is totally safe and pure are false.
No one should assume that just because water comes from a bottle that it is necessarily any purer or safer than most tap water. Testing commissioned by NRDC and studies by previous investigators show that bottled water is sometimes contaminated. NRDC contracted with three leading independent laboratories to do "snapshot" testing (testing one to three times for a subset of contaminants of concern) of bottled water.
We found after testing more than 1,000 bottles that about one fourth of the bottled water brands (23 of 103 waters, or 22 percent) were contaminated at levels violating strict enforceable state (California) limits for the state in which they were purchased, in at least one sample. We also found that almost one fifth of the waters we tested (18 of 103, or 17 percent) exceeded unenforceable sanitary guidelines for microbiological purity (heterotrophic-plate-count [HPC] bacteria guidelines, adopted in some states, the European Union (EU), and recommended by the bottled water industry) in at least one test. While HPC bacteria may be harmless themselves, they may mask the presence of pathogens; some states, the EU and the bottled water industry have adopted HPC guidelines to help ensure sanitary source water, processing, and bottling practices. In all, at least one sample of one third of the waters we tested (34 of 103, or 33 percent) exceeded a state enforceable standard for bacterial or chemical contamination, a nonenforceable microbiological-purity (HPC) guideline, or both.
The labs contracted by NRDC detected contaminants of potential concern (either microbes or chemicals regulated in tap or bottled water) in at least one sample of about half of the bottled waters we tested, though in the majority of the waters no standards were exceeded. While state or industry standards and guidelines were violated in at least one test for about one fourth of the bottled waters, just four waters (4 percent) exceeded the weak federal standards. Of these four waters, two violated the FDA coliform-bacteria rule (coliforms are bacteria that can be harmless themselves but may indicate the presence of fecal contamination and disease-carrying organisms in the water) in one test. When we retested another lot of the same waters for coliform bacteria, however, both of these waters tested clean. In addition, two other waters violated the FDA standard for fluoride in two sequential tests of samples from different lots of these two waters.
While our testing is the most comprehensive publicly available independent testing of U.S. bottled water, it must be viewed as incomplete. Only about half of the drinking water contaminants regulated by FDA and EPA were tested, due to cost constraints. There are, conservatively, more than 700 brands selling bottled water in the United States, yet we tested only 103 waters. Additionally, we generally tested just one to three lots of each water, whereas often thousands or even millions of bottles may be produced annually by a single bottler, with the potential for periodic (and undetected) contamination problems. Testing by other investigators generally has been consistent with our results. For example, as is discussed in detail in the accompanying Technical Report (print report only), a major survey of microbiological contamination of domestic and imported bottled water sold in Canada published in 1998 yielded results very similar to NRDC's.  We were not able to test for Cryptosporidium in bottled water (nor did the Canadian investigators) because the current EPA method for Cryptosporidium monitoring requires the filtration of many gallons of water and analysis of the filter using a method feasible for bottlers prior to bottling the water, but this was logistically and financially infeasible for us to use on finished product sold at stores.
Bottled water recalls and other contamination incidents -- whether bacterial, industrial-chemical, algae, excessive-chlorine, or other contamination problems -- have sometimes been quietly dealt with by bottlers, generally with little or no public fanfare. In other cases, violations of bottled water standards have been allowed to go on for months without a recall or formal enforcement action. Although most of the bottled water on the market seems to be of good quality, some of these products are not as absolutely pure and pristine as many of their consumers may expect.
Comparing the data for bottled water quality with those for tap water is not straightforward. Far more monitoring data are publicly available for tap water than for bottled water. EPA requires frequent monitoring of tap water and makes available on its Web site national compliance data for all tap water systems.  Additionally, numerous surveys of tap water quality (beyond simple compliance data) are available for tap water quality,  whereas no such comprehensive data are available for bottled water. Thus, direct comparison of tap water quality versus bottled water quality is not possible based on comparable databases. However, EPA recently reported that in 1996, almost 10 percent of community tap water systems (serving 14 percent of the U.S. population) violated federal EPA tap water treatment or contaminant standards, and 28 percent of these tap water systems violated significant water quality monitoring or reporting requirements.  While these tap water system compliance data are plagued by underreporting and likely understate the extent of the problem somewhat,  without question they are based on a far larger database than is publicly available for bottled water. Moreover, according to available data, nearly half of the U.S. population served by tap water systems gets legally allowable but from a health standpoint potentially significant levels of contaminants such as cancer-causing trihalomethanes, radon, and/or arsenic in their tap water.  Thus, while there definitely are problems with a substantial minority of the nation's tap water systems, based on the limited data available there is little basis to conclude that just because water is purchased in a bottle it is necessarily any better than most tap water.
2. Bottled water contamination with microbes may raise public health issues, particularly for people who are immunocompromised.
Millions of Americans use bottled water as their primary source of drinking water. Some of these people are immunocompromised (such as people undergoing cancer chemotherapy, organ-transplant recipients, the chronically ill elderly, some infants whose immune systems are not fully developed, and people with AIDS) and use bottled water at the recommendation of public health officials or health care providers, who suggest that tap water use may be too risky.[1b] In some cases, officials also may urge the general public to use bottled water during a tap water contamination crisis.
As discussed in Chapter 3 and our attached Technical Report (print report only), NRDC's testing and other published and unpublished data indicate that while most bottled water apparently is of high quality in terms of microbiological purity, a substantial minority of it may not be. As noted there, a small percentage of the bottled water we tested (about 3 percent) sometimes contained coliform bacteria -- a possible indicator of contamination with pathogenic bacteria -- and nearly one fifth of the waters we tested contained heterotrophic-plate-count (HPC) bacteria at levels exceeding state and industry guidelines in at least one test. Some bottled waters contain bacteria (sometimes naturally occurring), including species of Pseudomonas and others, some of which may be a health concern for immunocompromised people. 
In cases where there is known tap water microbial contamination, or where an individual suffers from specific health problems such as a compromised immune system, tap water can be boiled for one minute to kill all microbes. In the alternative, certain types of bottled water may be a temporary solution. To be cautious, however, an immunocompromised person should buy bottled water only if it is from a protected source, and is subjected to EPA-CDC-recommended treatment to kill Cryptosporidium, the intestinal parasite that sickened over 400,000 people and killed over 100 in a 1993 Milwaukee tap water incident.  For example, to remove or kill Cryptosporidium, water must be treated with "absolute one micron" membrane filtration or reverse osmosis, adequately high levels of ozone disinfection, or distillation, at a minimum.
Thus, NRDC recommends that seriously immunocompromised people boil their tap water for one minute before using it for consumption or washing food. If they choose to buy bottled water, they should consider purchasing only certified "sterile" bottled water. Most bottled water has not been independently certified to meet either the EPA-CDC standards for killing Cryptosporidium or the definition of "sterile" water, so vulnerable people must be especially careful in selecting a drinking water supply.[1c]
3. Government bottled water regulations and programs have serious deficiencies.
Chapter 4 outlines in detail the gaping holes in federal regulatory controls for bottled water, and the trivial FDA resources dedicated to protecting bottled water. FDA estimates that one half of a full-time FDA staff person is dedicated to bottled water regulation, and fewer than one FDA staff-person equivalent is spent on assuring compliance with FDA bottled water rules.  An estimated 60 to 70 percent of the bottled water sold in the United States, according to FDA interpretations, is exempted from FDA's contamination limits and specific bottled water standards because it is bottled and sold in the same state.
Thus, under FDA's interpretation, the regulation of most bottled water is left to ill-equipped and understaffed state governments. Yet 43 of 50 states have the equivalent of fewer than a single staff person dedicated to regulating bottled water, according to our 1998 state survey. Four states have adopted no regulations at all for bottled water, and the majority of states have simply republished FDA's deficient rules. About 40 states say they regulate "intrastate" waters, but most have dedicated virtually no resources to doing so.
FDA's rules also exempt many forms of what most of us would consider "bottled water" from all of its specific water-testing and contamination standards. If the product is declared on the ingredient label simply as "water," "carbonated water," "disinfected water," "filtered water," "seltzer water," "sparkling water," or "soda water," it is not considered "bottled water" by FDA,  nor, as noted in Chapter 4, do most states regulate this water as bottled water. For these products, the specific FDA contamination standards and water quality testing requirements for bottled water are not applicable. No contamination monitoring is required, and only a vague narrative legal standard applies, stating that the water cannot be "adulterated" -- a term not specifically defined and, to date, apparently never enforced against any of these products by FDA. Therefore, the generalized FDA "good manufacturing practice" requirements applicable to these waters set no specific contamination standards. The same is true with most state regulations.
Even what FDA defines to be "bottled water" is exempt from many of the standards and testing requirements that apply to tap water. This appears to directly contradict the letter and the spirit of the Federal Food, Drug, and Cosmetic Act (FFDCA), which requires -- under a provision strengthened in 1996 -- that FDA's bottled water standards must be at least as stringent as tap water standards.  For example, EPA's rules clearly prohibit tap water from containing any confirmed E. coli or fecal coliform bacteria (bacteria that are indicators of possible fecal matter contamination often associated with waterborne disease).  FDA has no such prohibition for bottled water; instead, any type of coliform bacteria is allowed up to a certain level.  (See Table 1 for a comparison of EPA and FDA rules.)
Similarly, a big city has to test its tap water 100 times or more each month for coliform bacteria -- many times a day, on average -- yet bottled water (even at an enormous bottling plant) must be tested for coliform bacteria only once a week under FDA rules. Moreover, while high overall levels of bacteria (known as heterotrophic-plate-count [HPC] bacteria) can be counted toward bacteria violations for city tap water (in the absence of adequate disinfection), as described in Chapter 4, FDA bowed to bottled water industry arguments and decided to apply no standards for HPC bacteria in bottled water. HPC bacteria are commonly found in bottled water.
EPA's "information collection rule" generally requires big cities that use surface water (such as rivers or lakes) for tap water to test for common parasites such as viruses, Giardia, and Cryptosporidium. Under FDA rules, water bottlers are never required to do so. In the same vein, cities using surface water generally must disinfect their water and filter it to remove bacteria and certain parasites.[1d] Yet there are no FDA standards requiring bottled water to be disinfected or treated in any way to remove bacteria or parasites. Additionally, the FDA requirement that bottled water be derived from an "approved source" is no substitute for source water protection, filtration, or disinfection. This rule has been aptly characterized as a "regulatory mirage," since what is "approved" is left to state discretion with no meaningful federal requirements or oversight.
For chemical contaminants, the regulations for bottled water are also weak in many ways. While a city generally must test its tap water for scores of organic chemicals (such as industrial chemicals, some pesticides, and trihalomethanes) at least quarterly,[1e] bottlers generally need only test once a year under FDA's rules. These infrequent annual tests could miss serious problems, because levels of these contaminants sometimes vary substantially depending on when they are tested.
Also, phthalate[1f] -- a toxic chemical produced in plastic-making that tests show can leach from plastic into water under common conditions -- is regulated by EPA in tap water but FDA does not regulate it in bottled water. After some water bottlers and plastics manufacturers argued that phthalate controls would be inappropriate and burdensome for bottled water, FDA decided not to regulate it in bottled water, where it is sometimes found, particularly after long storage.
Furthermore, FDA currently has no enforceable standard or treatment requirement for three other contaminants regulated by EPA in tap water -- acrylamide, asbestos, and epichlorohydrin. Thus, while city water systems generally must test for all of these contaminants and must meet EPA standards for them, presently water bottlers need not.
EPA also requires city tap water suppliers to test for more than a dozen "unregulated" contaminants -- chemicals that are not currently subject to EPA standards but which, if present, may pose a health concern, such as a risk of cancer. Under EPA rules, states are to consider adding 15 additional named unregulated contaminants to this list for mandatory water system monitoring, if they are believed to be a potential problem in local tap water.  Bottlers face no monitoring requirements for any unregulated contaminants.
Even if bottled water is more contaminated than FDA's standards would otherwise allow, FDA rules explicitly allow the water to be sold, as long as it says on the label "contains excessive chemical substances" or "contains excessive bacteria" or includes a similar statement on the label. FDA says it may enforce against such labeled contaminated water if it finds that it is "adulterated" and "injurious to health." However, there is no requirement that water bottlers report such problems to FDA, and apparently there are no cases of FDA having taken any enforcement action against any such bottlers.
FDA has stated that bottled water regulation carries a low priority.  Because of this, water bottlers can expect to be FDA-inspected only about every four to five years, on average.  This is far too infrequent to detect certain possible problems, such as periodic contamination caused by occasional substandard plant operations or maintenance, bacteria from sewage overflows or leaks, pest infestations, or occasional spikes of pollution due to short-lived phenomena. In addition, bottlers are not required to keep records of their operations and testing for more than two years, making effective inspections difficult or impossible, since evidence of periodic or past problems can simply be discarded before it is ever reviewed by inspectors.
It also should be noted than in many cases FDA's rules are weaker than international standards. The European Union's (EU's) bottled water standards, for example, set limits for total bacteria count,  which, as noted above, FDA does not. Moreover, the EU's bottled mineral water rules ban all parasites and pathogenic microorganisms, E. coli or other coliform bacteria, fecal streptococci (e.g., Streptococcus faecalis, recently renamed Enterococcus faecalis), Pseudomonas aeruginosa, or sporulated sulphite-reducing anaerobes, whereas FDA's rules include no such bans.  Additionally, unlike the FDA rules, EU rules require natural mineral water's labels to state the waters' "analytical composition, giving its characteristic constituents" and the specific water source and name, and information on certain treatments used.  The EU mineral water rules further forbid use of more than one brand label per source of water and generally prohibit labels from making any claims about the prevention, treatment or cure of human illness.  No such provisions are included in FDA rules. Similarly, the EU's new general standards for all bottled water generally are far stricter than FDA's rules, and FDA's standards for certain chemicals (such as arsenic) are weaker than World Health Organization (WHO) guidelines for drinking water. 
4. Voluntary bottled water industry controls are commendable, but an inadequate substitute for strong government rules and programs.
The bottled water industry's trade association, the International Bottled Water Association (IBWA), has sometimes been a progressive force in seeking to improve certain FDA controls (petitioning for stronger FDA rules in some areas, for example). Moreover, IBWA has adopted a voluntary state bottled water code -- somewhat stricter than the FDA rules -- which has been adopted in whole or in part by 16 states. However, IBWA sometimes has vigorously fought against tough FDA rules, such as possible controls on Pseudomonas aeruginosa bacteria, rules for heterotrophic bacteria, and right-to-know requirements for bottled water. The fight against right-to-know for bottled water is interesting in light of the bottled water industry's frequent references to tap water contamination problems. It also starkly contrasts with IBWA's admission that bottled water sales may have increased due to the requirement that diet soda labels disclose all ingredients, which IBWA said may have driven consumers concerned about diet soda's contents to use bottled water. 
IBWA has adopted a much-ballyhooed voluntary industry code and inspection program for its members. The association claims its members produce 85 percent of the bottled water sold in the United States.  But these voluntary IBWA standards are just that -- voluntary -- in the 34 states that have not adopted them, and there is no published reporting about compliance. Additionally, IBWA does not disclose the results of its inspections and testing to the public, so it is impossible to verify independently the effectiveness of these voluntary programs. Moreover, even by IBWA's count, many bottlers are not IBWA members and have never volunteered to comply with the association's standards. In fact, some of the problems with some bottled waters discussed in this report have occurred with IBWA members, suggesting the IBWA program is not foolproof. Finally, it should be noted that, as with FDA rules, IBWA standards do not apply to seltzer, soda water, carbonated water, or the many other waters exempt from FDA's bottled water rules. 
5. Bottled water marketing can be misleading.
Chapter 5 shows that despite recent FDA rules intended to reduce misleading marketing, some bottled water comes from sources that are vastly different from what the labels might lead consumers to believe. One brand of water discussed in this report was sold as "spring water" and its label showed a lake and mountains in the background -- with FDA's explicit blessing. But until recently the water actually came from a periodically contaminated well in an industrial facility's parking lot, near a waste dump (a state whistleblower informed the local media after years of internal struggles, finally putting an end to the use of this source). Another brand of water sold with a label stating it is "pure glacier water" actually came from a public water supply, according to state records. While FDA recently adopted rules intended to curb such practices, those rules include many weak spots and loopholes (including those that allowed the water taken from an industrial-park well to be sold as spring water with a label picturing mountains), and there are very few resources to enforce them.
Water with one brand name can come from numerous different sources, depending upon the time of year, location of sale, or other market factors. Moreover, water from one source (such as the industrial-parking-lot well noted above) can be used and labeled for a half-dozen or more different labels and brands. In addition, according to government and industry estimates, about one fourth or more of the bottled water sold in the United States  (and by some accounts 40 percent) is taken from public water systems -- tap water, essentially. Sometimes this tap water is bottled after additional treatment (such as carbon filtration or ozonation), and sometimes it is bottled with little or no additional treatment.
6. The long-term solution to drinking water problems is to fix tap water -- not to switch to bottled water.
Many people may choose to use bottled water because they prefer its taste and smell, or because it is convenient. Bottled water, in some cases, also may be needed as a stopgap measure when tap water is contaminated, rendering the water nonpotable (as in the case of a boil-water alert). In the long run, however, it is far better from an economic, environmental, and public health point of view to improve public drinking water supplies than it is to have a massive societal shift from consumer use of tap water to use of bottled water. We cannot give up on tap water safety. The reasons we have reached this conclusion include:
- Public health concerns. Bottled water sometimes poses its own potential health risks due to contamination. Furthermore, even if bottled water is completely pure, use of it can only somewhat reduce public exposure to contaminants in tap water; some people will continue to use tap water. Even if no one were to drink tap water, virtually everyone would continue to be exposed to some common contaminants (especially those that are volatile or can penetrate the skin) when showering, bathing, washing dishes, and cooking.
- Equity concerns. If those who can afford bottled water shift to it as their primary source of drinking water, only low-income people are left drinking tap water, its quality may then slip into an ever-downward spiral.
- Environmental concerns. Provision of water by underground pipe is energy-efficient and consumes far fewer natural resources per gallon than using bottled water. Placing water in bottles and transporting those heavy bottles around the country (or around the globe) consumes far more energy and other resources than using tap water. The manufacture of bottles also can cause release of phthalates, and other byproducts of plastic-making, into water, air, or other parts of the environment. And, ultimately, many bottles will be added to already overflowing landfills or incinerated, potentially adding to our environmental problems.
- Economic concerns. Bottled water typically costs hundreds of times more than tap water, even up to 10,000 or more times more than what comes out of your faucet. These costs cannot be easily borne by low-income people and should not have to be borne by the elderly, the immunocompromised, or chronically ill people in order to get water that is safe to drink. The $4 billion a year now spent by consumers on bottled water could be better spent on upgrading tap water supplies.
Thus, in NRDC's view, although bottled water may be a convenience or needed as a short-term solution to tap water contamination problems in some communities or for highly vulnerable subpopulations, it should generally be viewed only as a temporary fix. Our study leads us to make the following recommendations:
1. Fix tap water quality -- don't give up and just rely on bottled water.
For the reasons just noted, it would generally be better to upgrade and improve tap water quality than to have a part of society shift to bottled water. Those who dislike the taste and smell of their tap water may want to consider placing tap water in a glass or ceramic pitcher in their refrigerator, with the top loose to allow the chlorine to dissipate overnight. This also will allow volatile disinfection by-products to evaporate (though less volatile disinfection by-products may stay in the water). Overnight refrigeration in a loosely capped container eliminates the objectionable chlorine taste and odor, and the chilled water can be put in reusable sports bottles as desired to make it convenient to carry ice-cold water to the office, on trips, or when exercising. It also saves money and has environmental and other benefits, as previously noted.
2. Establish the public's right to know for bottled water as now required for tap water.
Bottled water labels should be required to list any contaminants found in the water (as well as health goals and standards), the water's fluoride and sodium content, the health effects of the contaminants found, the bottler's compliance with applicable standards, the source of the water, and any treatment used. Labels also should indicate whether the water meets the EPA-CDC criteria for Cryptosporidium safety. The date of bottling and information on how to get further information also should be placed on labels. We fail to understand why, if bottled water is as pure as the bottlers say, they are so afraid of a right-to-know requirement. However, FDA has the authority to require such information on bottled water labels, has been required by the Safe Drinking Water Act to evaluate the feasibility of doing so, and therefore should move forward with rules requiring such disclosure for bottled water.
3. FDA should create a Web site and a phone-accessible information system on bottled water.
FDA should add to its Web site and should make available, through a hot line, a user-friendly array of information on bottled water brands, including all of the basic information noted in recommendation 2, for each bottler. This bottled water information should build upon and expand the EPA hotline and web site that gives specific information on individual tap water systems and drinking water generally. The FDA hot line and Web site should make available the results of all government, industry, or other bottled water testing by certified labs for all brands. It also should include information on all inspections and recalls, and any other relevant consumer information on particular brands of bottled water.
4. Overhaul FDA rules for bottled water.
The FDA rules for bottled water are weak and should be strengthened. If necessary, FDA should request additional legislative authority to adopt these changes. FDA should:
- Establish standards and monitoring requirements for bottled water no less stringent than EPA's rules for tap water in major cities, including standards for all microbiological and chemical contaminants, specific and defined water treatment (including filtration and disinfection or strict source-protection requirements), operator-certification requirements, and unregulated-contaminant monitoring rules.
- Set strict, up-to-date standards for contaminants potentially found in bottled water. These standards should be at least as protective of public health as the strictest regulations adopted by other authorities. Thus, the standards should be as stringent as possible for the bottled water industry and certainly should be no less stringent than the following: arsenic less than 5 parts per billion (ppb)(California Proposition 65); heterotrophic-plate-count bacteria less than 100 colony-forming units per milliliter at bottling (EU standard), 200 cfu/ml 5 days after bottling in 90 percent of samples (industry recommendation), and a maximum at all times of 500 cfu/ml; no parasites, pathogens, fecal streptococci (e.g., the recently renamed Enterococcus faecalis), Pseudomonas aeruginosa, sporulated sulphite-reducing anaerobes (EU natural mineral water rules); trihalomethanes less than 10 ppb (California law and industry model code); phthalate less than 6 ppb (EPA tap water); individual synthetic organic and inorganic chemicals (e.g., bromodichloromethane) equal to California's Proposition 65 levels. For other contaminants more strictly controlled under bottled water industry code than under current FDA rules or with EPA Health Advisories, FDA should adopt the industry or EPA recommendation.
- Immediately finalize its 1993 proposed ban on coliform bacteria in bottled water.
- Establish clearly defined criteria and protections for an "approved source" of bottled water under FDA rules, and require annual state reevaluation of compliance with these new "approved source" rules, including review of potential contamination problems.
- Require bottlers to retain microbial test results for 5 years, and chemical tests for 10 years, as EPA requires for tap water.
- Mandate a bottling date and "refrigerate after opening" statement on labels, in order to inform consumers who seek to minimize the chances of potentially excessive microbial growth and contamination in bottled water.
- Require labs used for bottled water analysis to be certified by EPA or FDA.
- Direct that water be tested daily at the plant for microbes, quarterly for chemicals during bottling, and quarterly in bottles after extended storage, especially for chemicals that can leach from bottles and for microbes that can multiply during storage.
- Require quarterly reporting of test results to states and FDA, and reporting of acute violations within 24 hours to state and FDA officials.
- Prohibit all sales of water contaminated at levels above FDA standards.
- Apply FDA's standards to all intrastate bottled water sales.
- Mandate that water bottlers be trained and certified.
- Require state bottled water programs to be reviewed and approved by FDA, and FDA should oversee their effectiveness.
- Establish clear mandatory recall authority for FDA through administrative order or a civil action.
- Maintain an inventory, and register all water bottlers.
- Cover all water sold in a bottle that is likely to be ingested by people, including "purified," "disinfected," "seltzer," etc., under the FDA bottled water standards -- as under California and other states' laws.
- Conduct routine FDA monitoring of bottled water quality for waters sold across the country, as has been done in Canada for many years, and release the results, including brand names, to the public in published reports and on its website.
5. Annual inspections should be required.
FDA should conduct annual inspections (or fund annual state inspections) of all bottling facilities and of their water sources.
6. Institute a "penny-per-bottle" fee to assure bottled water safety.
We recommend that a fee of one cent per bottle of bottled water sold should be instituted, to be placed in a trust fund for use without further appropriation by FDA to pay for a stringent bottled water regulatory program. The fee, which we estimate would raise more than $30 million dollars a year, should fund improved FDA implementation, random testing, a public Web site, state and federal inspections, and funding and oversight of state programs and bottlers.
7. Set a deadline for transferring the bottled water program to EPA if FDA lacks the resources or will to implement it effectively.
FDA has made it clear that bottled water protection is a low priority. If FDA concludes that making bottled water comply with the same requirements as tap water is unduly burdensome, or that the preceding recommendations to achieve that goal are not of sufficient priority to claim FDA resources, the program should be transferred to EPA, which already regulates tap water. FDA should be given no more than 18 months to demonstrate, by overhauling its rules and program, whether it wishes to retain the program. If such an overhaul does not occur, the program should be automatically transferred to EPA. EPA should be given six months to apply the rules applicable to big city water systems to bottled water; of course, the rules should be modified where they would be inapplicable to bottled water (as where EPA rules require monitoring at the tap). EPA also should be provided the revenue from a penny-per-bottle fee on bottled water to carry out the program. We make this recommendation for transfer with some uneasiness, since EPA's tap water regulatory program suffers from its own serious deficiencies and resource constraints. However, on balance we believe that if FDA continues to lack the will and resources to address bottled water issues as the sales skyrocket, even an inadequate EPA bottled water regulatory program could hardly be worse than FDA's current effort.
8. Establish "certified safe" bottled water.
In light of the poor government regulatory performance, an independent third-party organization such as Green Seal or Underwriters Labs should establish a "certified safe" bottled water program. Criteria for inclusion would be that the water always meets the strictest of all standards, including FDA, IBWA, international (e.g., EU and WHO) and state rules, recommendations, and guidelines, meets all EPA health goals, health advisories, and national primary drinking water regulations, is tested at least daily for microbial contaminants and quarterly for chemicals (monthly if using surface water or other water subject to frequent water quality changes), meets source-water protection criteria, is protected from Cryptosporidium in accordance with EPA-CDC guidelines, is disinfected, and is surprise inspected twice a year by independent third-party inspectors. The certifying organization should establish an open-docket release of its inspection, testing, and compliance evaluation results. While the current NSF and IBWA seals are intended to provide such a stamp of approval, we believe a more independent and open body imposing stricter standards and making all testing, inspection, and other collected information readily available to consumers (including on the Web), would provide greater consumer confidence in the certification.
Thus, we believe the long-term national solution is to fix the nation's tap water supplies. Until the recommended regulatory changes are adopted, those who wish to use bottled water for reasons of taste or otherwise cannot be confident that they are necessarily getting what they pay for -- a pure, well-regulated product. Unless such reforms are adopted, bottled water consumers should observe the ancient rule of caveat emptor -- "buyer beware.")
1a. Throughout this document we use the term contaminant in the same way that term is used in the Safe Drinking Water Act (SDWA) -- i.e, "any physical, chemical, biological, or radiological substance or matter in water." 42 U.S.C. § 300f(6).
1b. EPA and CDC have jointly recommended that severely immunocompromised people consult with their health care provider to decide whether they should drink tap water or switch to bottled water treated with certain advanced technologies (or use tap water that is boiled or treated with an advanced home filter). However, we have found that very few bottled water companies clearly label their bottles to enable consumers to determine whether the water meets the EPA-CDC recommendations.
1c. The use of home filtration devices is an issue beyond the scope of this study, but experts recommend that at a minimum, an immunocompromised person should only purchase a filter certified by NSF International for "cyst removal" (i.e., to remove protozoa "cysts," such as Cryptosporidium). In addition, users of home filters must be extremely careful to maintain the filter and to change the filtration media at least as frequently as recommended by the manufacturer, or more often.
1d. Cities using surface water as their source generally must disinfect, unless they can document and obtain state approval for a filtration waiver, based on evidence that their source water is pure and highly protected from contamination.
1e. In certain cases, EPA's rules allow tap water to be tested less frequently than quarterly for some organic contaminants. For example, a waiver may be available to a system if the contaminant was not detected in the first round of four quarterly tests and the system is evaluated by the state and found unlikely to become contaminated in the future.
1f. Specifically, di(2-ethylhexyl)phthalate, or DEHP--a likely carcinogen that studies have indicated also may cause disruption of the endocrine system. See, e.g., B.J. Davis, R.R. Maronpot, and J.J. Heindel, " Di-(2-ethylhexyl) phthalate Suppresses Estradiol and Ovulation in Cycling Rats," Toxicol Appl Pharmacol, vol. 128, no. 2, pp. 216-223 (October 1994),(exposure to DEHP resulted in hypoestrogenic anovulatory cycles and polycystic ovaries in adult female rats).
2. The bottled water NRDC purchased ranged in price from a low of about $0.70 per gallon to more than $5.00 per gallon for more expensive imports sold in smaller bottles. The average cost of bottled water in California has been reported to be $0.90 cents per gallon, though that appears to be a low estimate compared to most of our purchases. Tap water generally costs from a low of around $0.45 cents per thousand gallons to about $2.85 per thousand gallons, with an average cost of about $1.60. L. Allen & J.L. Darby, "Quality Control of Bottled and Vended Water in California: A Review and Comparison of Tap Water," Journal of Environmental Health, vol. 56, no. 8, pp. 17-22 (April 1994); "Bottled Water Regulation," Hearing of the Subcommittee on Oversight and Investigations of the House Committee on Energy and Commerce, Serial No. 102-36, 102nd Cong., 1st Sess. 5, (April 10, 1991). Thus, the ratio for bottled water to tap water ranges from a low of about 240 times more expensive (cheap bottled water: expensive tap water), to over 10,000 times more expensive (expensive bottled water: cheap tap water).
4. D. Warburton, B. Harrison, C. Crawford, R. Foster, C. Fox, L. Gour, and P. Krol, "A Further Review of the Microbiological Quality of Bottled Water Sold in Canada: 1992-1997 Survey Results," International Journal of Food Microbiology, vol. 39, pp. 221-226 (1998).
6. See, e.g. NRDC, USPIRG, and Clean Water Action, Trouble on Tap: Arsenic, Radon, and Trihalomethanes in Our Drinking Water (1995); NRDC, Your Are What You Drink (1995); NRDC, Think Before You Drink (1993); NRDC, Think Before You Drink: Urgent Release: 1992-1993 Update (1994); EWG & NRDC, Just Add Water (1996).
7. EPA, Providing Safe Drinking Water in America: 1996 National Public Water System Annual Compliance Report and Update on Implementation of the 1996 Safe Drinking Water Act Amendments, Executive Summary (September 1998),(available at www.epa.gov/ogwdw).
10. See, e.g., D.W. Warburton, "A Review of the Microbiological Quality of Bottled Water Sold in Canada, Part 2: The Need for More Stringent Standards and Regulations," Canadian J. of Microbiology, vol. 39, p. 162 (1993); H. Hernandez-Duquino, and F.A. Rosenberg, "Antibiotic-Resistant Pseudomonas in Bottled Drinking Water," Canadian J. of Microbiology, vol. 33, pp. 286-289 (1987); P.R. Hunter, "The Microbiology of Bottled Natural Mineral Waters," J. Applied Bacteriol., vol. 74, pp. 345-352 (1993); see also, F.A. Rosenberg, "The Bacterial Flora of Bottled Waters and Potential Problems Associated With the Presence of Antibiotic-Resistant Species," in Proceedings of the Bottled Water Workshop, September 13 and 14, 1990, A Report Prepared for the Use of the Subcommittee on Oversight and Investigations of the Committee on Energy and Commerce, U.S. House of Representatives, Committee Print 101-X, 101st Cong., 2d Sess. pp. 72-83 (December 1990).
11. See, e.g. , W. R. MacKenzie, et al., "A Massive Outbreak in Milwaukee of Cryptosporidium Infection Transmitted Through the Public Water Supply," New Engl. J. of Med. vol. 331, no. 3, pp. 161-167 (July 21, 1994); Marilyn Marchione, "Silent Disaster: Crypto Has Killed 104 -- And Counting," Milwaukee Journal, p. 1 (March 27, 1994).
21. European Union, Council Directive of 15 July 1980 on the Approximation of the Laws of the Member States Relating to the Exploitation and Marketing of Natural Mineral Waters, Article 5 § 1 (80/777/EEC: OJ No. L 229, 30.8. 1980 p. 1), as amended (available in consolidated text form at www.europa.eu.int);EU, Council Directive 98/83/EC of 3 November 1998 On The Quality of Water Intended for Human Consumption [available at same web site].
26. EU, Council Directive 98/83/EC, Supra; The WHO provisional guideline for arsenic in drinking water for human consumption is 10 ppb. World Health Organization, Guidelines for Drinking Water Quality (2nd Edition, Geneva, 1993). The FDA standard for arsenic (and the EPA tap water standard, required to be updated in 2001), based on an outdated 1942 U.S. Public Health Service guideline, is 50 ppb.
30. This troubling case arose in Massachusetts. Massachusetts state files reveal that the described well in Millis, Massachusetts for years supplied several bottlers, including Cumberland Farms, West Lynn Creamery, Garelick Farms, and Spring Hill Dairy for sale as "spring water" under many brand names. Massachusetts Department of Public Health, Ann & Hope Water Incident Files, 1993-1997; MDPH Memoranda Provided to NRDC Pursuant to Freedom of Information Request; Personal Communication with Dr. Liz Bourque, MDPH, August 1997.
31. According to State of Washington files, bottled water called "Alaska Premium Glacier Drinking Water: Pure Glacier Water From the Last Unpolluted Frontier, Bacteria Free" actually was derived from"Public Water System #111241," apparently a public water system in (Juneau, Alaska), according to the files. The bottler apparently was told that when it reordered its labels, it had to state that the water is "from a municipal source" or "from a community water system" in accordance with FDA rules; the phrase "Pure Glacier Water" was, according to State files, "considered false and misleading." Also, the bottler was required to drop the "bacteria free" claim, as this was "considered synonymous with sterile and false." Washington State Department of Agriculture Food Establishment Inspection Report 4/17/97 and attachments; WSDA Food Establishment Inspection Report 10/4/96 and attachments; WSDA Food Processor Licensing Worksheet and Attachments, and WSDA Food Establishment Inspection Report and Attachments, 3/20/96. State officials report that the required label changes have been made after the intervention of FDA and state regulators. Personal communication with Shelly Haywood, USDA (January 1999)
32. L. Allen & J.L. Darby, "Quality Control of Bottled and Vended Water in California: A Review and Comparison of Tap Water," Journal of Environmental Health, Vol. 56, No. 8, p. 19 (April 1994), citing FDA; accord, "Bottled Water Regulation," Hearing Before the Subcommittee on Oversight and Investigation of the House Committee on Energy and Commerce, 102nd Cong., 1st Sess., p. 3, Serial No. 102-36 (April 10, 1991); accord, Ibid. at 152 (Statement of William F. Deal, CEO, International Bottled Water Association). In a recent interview with the head of the FDA bottled water program, FDA confirmed that they have no reason to believe that this percentage has changed substantially since 1991. Interview with Terry Troxel, FDA, September 18, 1997.
33. Memorandum, Dr. Karen Golden, FDA:CFSAN:OC:RCS, Regarding Discussion with Tyrone Wilson, International Bottled Water Association, Regarding Bottled Drinking Water (dated February 10, 1992)[FDA Docket 93N-0200, Reference 2].
Bottled Water : Pure Drink or Pure Hype?. By Erik D. Olson. April 1999. Print version, $14.00. Order print copies .
last revised 7/15/2013
Get Updates and Alerts
Water on Switchboard
NRDC experts write about water efficiency, green infrastructure and climate on the NRDC blog.
Recent Water Posts
- After South Carolina Floods, We Must Act to Prevent Similar Disasters
- posted by Ben Chou, 10/8/15
- There's Still Time for the California Energy Commission to Fix the Drought Rebate Program
- posted by Ben Chou, 10/7/15
- Let's Be Honest About Environmental Safeguards
- posted by Jon Devine, 10/6/15
NRDC Gets Top Ratings from the Charity Watchdogs
- Charity Navigator awards NRDC its 4-star top rating.
- Worth magazine named NRDC one of America's 100 best charities.
- NRDC meets the highest standards of the Wise Giving Alliance of the Better Business Bureau.