header graphic


News Release (10/04/06)
EWG Analysis (10/04/06)
CDC Study (pdf)


PDF | CHPAC Letter to EPA

PDF | EPA Response to CHPAC



Rocket Fuel Contamination in California Milk

Suspect Salads

Rocket Fuel in Lettuce

Rocket Science

Rocket Fuel in Drinking Water

Human Testing of Perchlorate is Unethical


44 Million Women at Risk of Thyroid Deficiency From Rocket Fuel Chemical

Federal Study Confirms Perchlorate as
Widespread Public Health Threat

Although regulators have known for years that the rocket fuel chemical perchlorate contaminates hundreds of drinking water supplies across the country, new scientific evidence clearly shows that perchlorate is a much greater public health threat than previously realized. Tests of almost 3,000 human urine and breast milk samples — along with tests of more than 1,000 fruit, vegetable, cow's milk, beer, and wine samples — reveal that perchlorate exposure in the population is pervasive. And a startling new Centers for Disease Control study (CDC), released today, shows that perchlorate exposure is related to reduced thyroid hormone levels in women, particularly those with lower iodide intake. An Environmental Working Group (EWG) analysis shows that 44 million women are at particular risk to perchlorate-related health effects.

To deal with this serious public health issue, EWG urges the federal government to act promptly and set a drinking water standard of no more than 0.1 ppb of perchlorate. With new evidence showing widespread food contamination and effects on the thyroid at typical exposure levels, perchlorate exposure through drinking water cannot be tolerated. EWG also urges the Food and Drug Administration (FDA) to consider making the iodization of salt mandatory, given that not getting enough iodide in one's diet can compound perchlorate's health effects and iodide deficiency has increased sharply since the 1970s.

The vast majority of perchlorate manufactured in the U.S. is used by the Department of Defense to make solid rocket and missile fuel, while smaller amounts of perchlorate are also used to make firework and road flares. Perchlorate is also a contaminant of certain types of fertilizer that were widely used in the early part of the 20th century but are in limited use today. [Dasgupta 2006] According to EWG analysis of the Environmental Protection Agency's (EPA) latest data, perchlorate is known to be contaminating at least 160 public drinking water systems in 26 states. [EPA 2005]

In July, Massachusetts set the nation's first drinking water standard for perchlorate of 2 parts per billion (ppb), while California and New Jersey are currently considering standards of 6 ppb and 5 ppb, respectively. [Stateside 2006] The EPA has yet to set a federal drinking water standard, but has issued a controversial cleanup "guidance" level of 24.5 ppb. In March, a federal advisory committee on children's health sharply criticized the agency, writing in a letter to EPA Administrator Stephen Johnson that the guidance was "not supported by the underlying science and can result in exposures that pose neurodevelopmental risks in early life." [CHPAC 2006]

Perchlorate contamination is widespread in food, water, and people

In 2003, EWG and the Riverside Press Enterprise independently tested store-bought winter lettuce for perchlorate and found the contaminant in more than half of the samples tested — in some cases at high levels. [EWG 2003, Danelski 2003] A year later EWG tested California milk for perchlorate, finding the chemical in 31 out of 32 samples. [EWG 2004] Around the same time, the California Department of Food and Agriculture secretly conducted their own milk tests and found perchlorate in all 32 samples collected. [CDFA 2004] These studies were some of the first indications that food might be an important route for perchlorate exposure in addition to contaminated drinking water.

Only a couple of years, but more than 1,000 tests by government and independent scientists later, there can be no debate: The US population is being widely exposed to perchlorate via the food supply.

Perchlorate is a common contaminant of the food supply

Test results from 1090 food and beverage samples.

Type of food tested Number of samples tested Average perchlorate level (g/L or g/kg) Maximum perchlorate level (g/L or g/kg) Percent of samples with detectable perchlorate* Sample origin
Fruit 40 29.9 464 100% US + 20 other countries
Vegetables (other than lettuce) 81 28.8 628 41% 7 US states + Canada
Lettuce 526 14.6 370 47% 10 US states + Canada
Wine 77 5.0 50.3 100% US + 20 other countries
Milk 222 4.3 11.3 98% 22 US states
Beer 144 1.0 21.1 100% US + 47 other countries

*Note: The detection limits of different studies vary widely; for this reason, the percentages listed here may not reflect the actual frequency of perchlorate contamination in different foods. Lettuce and vegetable percentages are estimates due to incomplete data availability.

References: [CDFA 2004, Danelski and Beeman 2003, El Aribi et al. 2006, EWG 2003, EWG 2004, FDA 2004, Kirk et al. 2003, Kirk et al. 2005, Sanchez et al. 2005]

Perchlorate has been found in a wide variety of domestic and imported produce, with some of the highest levels being found in oranges, grapes, raspberries, apricots, melons, lettuce, tomatoes, basil, kale, spinach, and asparagus, among others. [El Aribi et al. 2006, Sanchez et al 2005, FDA 2004, EWG 2003, Danelski 2003] The chemical has also been found in 98 percent of 222 milk samples collected from 22 US states, and in hundreds of samples of beer and wine. [EWG 2004, CDFA 2004, Kirk et al. 2003, FDA 2004, Kirk et al. 2005, El Aribi et al. 2006] Fruits and vegetables had the highest concentrations, averaging about 30 micrograms of perchlorate per kilogram (g/kg) and ranging up to 628 g/kg in one US-grown spinach sample. Overall, 69 percent of the 1,090 food and beverage samples tested had detectable perchlorate.

Perchlorate is also polluting drinking water systems serving millions of Americans. According to tests conducted under the EPA's Unregulated Contaminant Monitoring Rule, at least 160 public water systems in 26 states are contaminated with perchlorate. [EPA 2005] And these figures don't count agricultural or private wells: In California's Santa Clara County alone, more than private 400 wells have been contaminated a flare manufacturing plant. [SCVWD 2004]

With this amount of perchlorate being found in the nation's food and drinking water it should come as no surprise that it is showing up in people's bodies as well. In 2005, scientists at Texas Tech University found perchlorate in every one of the 36 samples of breast milk they tested from women in 18 states. [Kirk et al, 2005] The average level found was 10.4 ppb, while the highest was 92 ppb. In September, the Centers for Disease Control (CDC) released new data from its National Health and Nutrition and Examination Survey (NHANES) program showing that all 2820 people tested had some amount of perchlorate in their urine. The average level found in this statistically representative sample of the US population was 5.5 ppb, and the highest was 160 ppb. [CDC 2006]

Perchlorate is found widely in people

Test results from 2856 urine and breast milk samples.

Human sample tested Number of samples tested Average perchlorate level (g/L, or ppb) Maximum perchlorate level* (g/L, or ppb) Percent of samples with detectable perchlorate Sample origin
Urine 2820 5.5 160 100% Statistical sample of the US population
Breast milk 36 10.4 92.2 100% 18 US states

*Note: To facilitate comparison with breast milk, urinary perchlorate values were not adjusted for urine concentration (which would require perchlorate to be expressed in g perchlorate per g of creatinine in urine).

References: [CDC 2006, Kirk et al. 2005]

The "safe dose" debate

Scientists and regulators have been engaged in a heated debate over what should be considered a safe dose of perchlorate for years. The chemical was first discovered to affect the thyroid in the 1950s and was even used medically for a time to lower thyroid hormone levels in patients with Grave's disease (severe hyperthyroidism). But until perchlorate started showing up at toxic waste sites and then in public drinking water supplies, no one had really looked into how perchlorate might affect the body at low levels. [EWG 2001]

Perchlorate acts by inhibiting the thyroid's ability to take up the nutrient iodide, which is a key building block for thyroid hormones. If the thyroid gland does not have enough iodide for a sufficient period of time, the body's thyroid hormone levels will eventually drop. Hypothyroidism (low thyroid hormone levels) in adults can cause fatigue, depression, anxiety, unexplained weight gain, hair loss and low libido. More serious, however, are the effects of thyroid hormone disruption in the developing fetus and child: Small changes in maternal thyroid hormone levels during pregnancy have been associated with reduced IQs in children. Fetuses, infants and children who experience more significant changes in hormone levels may suffer mental retardation, loss of hearing and speech, abnormal testicular development or deficits in motor skills. [EWG 2003]

In the early 1990s, the EPA began to conduct studies that involved feeding low doses of perchlorate to animals and looking for adverse effects. In 1995 the EPA found that laboratory animals developed thyroid disorders after two weeks of drinking perchlorate-laced water. In a 90-day drinking water study, researchers found significant reductions of thyroid hormone levels from perchlorate doses more than 10 times lower than those consumed in the two-week test. Subsequent studies found effects on brain and thyroid structure at even lower doses, and noted that rat pups born to exposed mothers were particularly like to show adverse effects. [EWG 2001, 2003]

Based on the findings of these animal studies, in 2002 the Environmental Protection Agency proposed a "reference dose," or safe dose level, for perchlorate equivalent to 1 ppb in drinking water for adults. This was a far cry from the 200 ppb level that the Department of Defense and its contractors claimed was safe, and the associated uproar eventually led to a National Academy of Sciences (NAS) review of the EPA's proposed reference dose. (Even this review became controversial after it was discovered that a paid consultant to the perchlorate polluter Lockheed Martin was appointed to the Academies' scientific committee; he was later asked to resign.) [Sass 2005]

One of the key elements of the safe dose debate was whether regulators should rely on the set of animal studies showing adverse effects at very low exposure levels, or whether they should rely on a small industry-sponsored human study now known simply as the "Greer study." Despite the fact that the study was extremely small (involving only 37 people), extremely short (perchlorate exposure was for only 14 days), and was conducted on a population of low concern (healthy, non-pregnant adults with high iodide intake), the NAS concluded that this study was more appropriate to use than the animal studies. [EWG 2003]

Based on its interpretation of the Greer study, the NAS recommended a safe dose level that is equivalent to 24.5 ppb perchlorate in drinking water — if one doesn't consider food exposure or the higher dose per bodyweight that infants and children get. Even this level, the NAS said, should be considered very conservative as it included a 10-fold safety factor to protect sensitive populations. Just one month later, in February of 2005, the EPA adopted the Academy's recommendation as its official reference dose. A year later, the EPA issued a guidance that set a preliminary remediation goal (PRG) for perchlorate in groundwater at Superfund sites of 24.5 ppb.

EPA's decisions have been widely criticized. In neither case did the agency provide an opportunity for public input, and the translation of its reference dose into a cleanup guidance pointedly ignored the significant exposures from food. The Children's Health Protection Advisory Committee (CHPAC) had some of the strongest criticisms of the agency. In its March 8th letter, the committee told EPA Administrator Stephen Johnson that the PRG is "not protective at children's health," " is not supported by the underlying science," "can lead to exposures that are well above USEPA'S [reference dose] for perchlorate," and "can result in exposures that pose neurodevelopmental risks in early life." [CHPAC 2006]

The committee went on to say that:

"CHPAC finds it disturbing that this change in the PRG was made without dissemination of a decision support document or any opportunity for public input. We recommend that [EPA] lower the PRG, taking into account infant exposures and susceptibility. We also recommend that USEPA's Office of Ground Water and Drinking Water develop a Maximum Contaminant Level for perchlorate, and in the interim, issue a health advisory for potable water that takes into account early life exposures."

The EPA responded to CHPAC's letter in May, but dismissed all of the committee's concerns. The agency claimed that the information on perchlorate in food was "too limited to characterize exposure to perchlorate on a national scale" so not considering food exposure at all was somehow the most "scientifically defensible approach." EPA also downplayed the committee's concerns about children's exposure by claiming that the EPA/NAS safe dose level was highly conservative, and implying that even exposures that were significantly higher than this level could not possibly cause any adverse effects in children or adults. [EPA 2006]

The EPA's comments are eerily reminiscent of those by the Council on Water Quality, an industry front group funded by industrial perchlorate polluters. On its website, which is subtitled "Facts about Perchlorate," the Council claims "there is no evidence that minute levels of perchlorate pose any health risk to anyone," and "a person would have to drink at least 500 gallons of water with 20 ppb of perchlorate in each gallon every day before there could be any adverse effect." [CWQ 2006]

Unfortunately, it appears that the EPA, the NAS, and the Council of Water Quality were all dead wrong.

New CDC study shows perchlorate affects people at every day exposure levels

Last February, the independent newsletter Risk Policy Report reported that the White House Office of Science & Technology Policy was pressuring the CDC to delay releasing data showing widespread perchlorate exposure among the US population. [EWG 2006] An EPA source told the newsletter that CDC had found levels of perchlorate that "leave no margin of safety" for the public, compared to EPA's current risk limit. The remark was remarkably prescient.

In a long-awaited study published in the journal Environmental Health Perspectives today, researchers from the Centers for Disease Control found that perchlorate exposure is affecting thyroid hormone levels in American women, particularly those with lower iodide intake. CDC researchers analyzed urine samples from more than 1,100 women for perchlorate, and then looked to see if perchlorate exposure could predict thyroid hormone levels. They found a relationship between perchlorate levels and one type of thyroid hormone (TSH) for all women. For those women with lower iodide intake (less than 100 micrograms of iodide per liter of urine), researchers found a relationship for two types of thyroid hormones (TSH and T4). Thirty-six percent of US women have iodide intake that fall into this category. [Blount et al. 2006, Tobin 2006]

The CDC study is highly significant because it is the first major epidemiological study to examine perchlorate-related health effects in the general population. In the Environmental Health Perspectives press release associated with the article, the reductions in thyroid hormones levels were noted as "significant" and "indicate that even small increases in perchlorate exposure may inhibit the thyroid's ability to absorb iodine from the bloodstream." [Tobin 2006]

The median urinary perchlorate level CDC's study was just 2.9 micrograms per liter. Since average urine output is about 1.5 liters per day, this means that women in the study were ingesting somewhere around 5 micrograms of perchlorate per day. Even at this low level, researchers found effects on thyroid hormone levels. But the EPA/NAS "safe dose" level corresponds to almost ten times this dose — 49 micrograms of perchlorate per day for a 70 kg adult. This clearly indicates that the current EPA reference dose and cleanup guidance is several orders of magnitude too high.

Women with low iodide intake are not the only population to be at particular risk for adverse effects related to perchlorate exposure. Also at particular risk are women who already have hypothyroidism (levels of thyroid hormones that are too low), and/or are pregnant. Pregnant women and their fetuses are of highest concern for several reasons: (1) pregnancy itself puts stress on the thyroid, and pregnancy-induced hypothyroidism is not uncommon; (2) the fetus is entirely reliant upon the mother for thyroid hormones during parts of gestation; (3) proper levels of thyroid hormones are necessary for brain and organ development, and even small disruptions can cause reduced IQ later in life.

EWG analysis of the CDC's most recent National Health and Nutrition Examination Survey (NHANES III), shows that approximately 44 million women have lower iodide intake, are hypothyroid, and/or are pregnant — and therefore at particular risk to perchlorate-related health effects.


Blount B, Pirkle JL, Osterloh JD, Valentin-Blasini L, Caldwell KL. 2006. Urinary Perchlorate and Thyroid Hormone Levels in Adolescent and Adult Men and Women Living in the United States. Environmental Health Perspectives. Online 5 October 2006. doi:10.1289/ehp.9466 Available at http://dx.doi.org/

California Department of Food and Agriculture. 2004. Data cited in: Sharp, R. 2004. Rocket fuel contamination in California milk. Environmental Working Group. Available at http://www.ewg.org/reports/rocketmilk/

Centers for Disease Control. 2006. NHANES 2001-2002 Surplus specimen data. Available at: http://www.cdc.gov/nchs/about/major/nhanes/lab01_02.htm

Children's Health Protection Advisory Committee. 2006. Letter to EPA Administrator Stephen Johnson. March 8, 2006. Available at: http://www.ewg.org/issues_content/perchlorate/20060315/chpac-epa_ltr.pdf

Council on Water Quality. 2006. "Facts about Perchlorate and Milk" and "Health Effects" fact sheets. Available at: http://www.councilonwaterquality.org/facts/milk.html and http://www.councilonwaterquality.org/science/health_effects.html

Danelski D., Beeman D. 2003. Special Report: Growing concerns: While scientists debate the risks, a study finds the rocket-fuel chemical in inland lettuce. The Press-Enterprise. April 27, 2003

Dasgupta PK, Dyke JV, Kirk AB, Jackson WA. 2006. Perchlorate in the United States. Analysis of Relative Source Contributions to the Food Chain. Environ. Sci. Technol. ASAP Web Release Date: 27-Sep-2006. DOI: Available at http://pubs3.acs.org/acs/journals/doilookup?in_doi=10.1021/es061321z

El Aribi, H., Le Blanc Y.J.C., Antonsen S., Sakuma T. 2006. Analysis of perchlorate in foods and beverages by ion chromatography coupled with tandem mass spectrometry (IC-ESI-MS/MS). Analytica Chimica Acta. 567(1): 39-47.

Environmental Protection Agency. 2005. Unregulated Contaminant Monitoring Rule data. Updated January 2005. Available at http://www.epa.gov/safewater/ucmr/data.html

Environmental Protection Agency. 2006. Letter from EPA Assistant Administrator Susan Parker Bodine to Melanie Marty, Chair of the Children's Health Protection Advisory Committee. May 11, 2006. Available at: http://yosemite.epa.gov/ochp/ochpweb.nsf/content/5112006.htm/

Environmental Working Group. 2001. Rocket Science: Perchlorate and the toxic legacy of the cold war. Available at: http://www.ewg.org/reports/rocketscience/es.html

Environmental Working Group. 2003. Rocket Fuel in Drinking Water: New Studies Show Harm From Much Lower Doses. Available at: http://www.ewg.org/reports/rocketwater/healtheffects.php

Environmental Working Group. 2003. Suspect Salads: Toxic rocket fuel found in samples of winter lettuce. Available at http://www.ewg.org/reports/suspectsalads/

Environmental Working Group. 2004. Rocket fuel contamination in California milk. Available at http://www.ewg.org/reports/rocketmilk/

Environmental Working Group. 2006. White House delays release of study showing toxic rocket fuel in most Americans. March 3, 2006. Available at: http://www.ewg.org/issues/perchlorate/20060303/index.php

Food and Drug Administration. 2004. Exploratory Data on Perchlorate in Food. Available at http://www.cfsan.fda.gov/~dms/clo4data.html

Kirk AB, Smith EE, Tian K, Anderson TA, Dasgupta PK. 2003. Perchlorate in milk. Environ Sci Technol. 37(21):4979-81.

Kirk AB, Martinelango PK, Tian K, Dutta A, Smith EE, Dasgupta PK. 2005. Perchlorate and Iodide in Dairy and Breast Milk. Environ Sci Technol. 39(7):2011.

Sanchez CA, Crump KS, Krieger RI, Khandaker NR, Gibbs JP. 2005. Perchlorate and nitrate in leafy vegetables of North America. Environ Sci Technol. 39(24):9391-7.

Santa Clara Valley Water District. 2004. $1.75 million aid for cleanup. Senate money to clear contamination in South County wells. Available at: http://www.valleywater.org/Water/Water_Quality/

Sass J and Solomon G. 2005. Inappropriate Influence by Industry on EHP News Article. Environ Health Perspect. 2005 February; 113(2): A87-A88.

Stateside Associates. 2006. States move ahead to develop perchlorate cleanup standards. State Environment Highlights. September 2006. Available at http://www.stateside.com/publications/environmental.shtml

Tobin J. 2006. New Study Suggests Perchlorate Effects on Thyroid Function of U.S. Women. Environmental Health Perspectives. Available at: http://www.ehponline.org/press/20061004.html

Support EWG - Donate Now