Sign up to receive email updates, action alerts & health tips from EWG. [Privacy]

2. Illegal Pesticides in Produce

Forbidden Fruit | Illegal Pesticides in the US Food Supply: 2. Illegal Pesticides in Produce

February 1, 1995

Forbidden Fruit analyzes 14,923 records of FDA pesticide monitoring data on 42 fruits and vegetables for the years 1992 and 1993. (For a detailed explanation of the methodology, see the Appendix.) During this two-year period, the FDA reported that 470 of the 14,923 samples, or 3.1 percent, contained illegal residues. Our analysis yielded 826 violations in the same 14,923 samples, for an overall violation rate of 5.6 percent (Figure 1).

The ubiquity of illegal pesticides in these fruits and vegetables seriously erodes the integrity of the pesticide food safety system (See The Health Effects of Pervasive Illegal Pesticides in Food). This would not be the case if violations were isolated occurrences. Unfortunately, they are not. Illegal pesticides are pervasive and systemic across the fruit and vegetable industry. In fact, as will be shown below, from one-third to one-half of the pesticides detected on five common foods over the two-year period were illegal.

We found five different types of pesticide violations among the data we reviewed (Table 2).

These include:

  • No-tolerance violations--pesticides which have no tolerance for the crops on which they are detected. Most of these residues represent an illegal use of the pesticide.
  • Over-tolerance violations--pesticides found at levels exceeding legal limits.
  • Action level violations--banned pesticides that persist in the environment found at levels exceeding FDA-established limits.
  • Trace no-tolerance violations--pesticides with no tolerance that are positively identified at levels too small to be quantified by FDA methods. These positively identified illegal residues represent an illegal use of a pesticide.
  • Canceled pesticides--pesticides banned for use on some or all crops in the U.S.

No-tolerance violations

The FDA's biggest failure is in recognizing "no-tolerance" violations--situations where the pesticide is not allowed for use on that crop (or in some cases on any crop), but where residues of the pesticide were identified and quantified by the FDA.

The FDA identified and reported 372 of these no-tolerance situations as violations; the Environmental Working Group found that the FDA had quantified, but failed to report as illegal, 183 additional "no-tolerance" violations, making 555 the true total of FDA-identified no-tolerance pesticides. Nearly all of these no-tolerance violations represent the illegal use of a pesticide. Some very small percentage may result from cross contamination of packing facilities that are tainted with post harvest pesticides applied to other crops.

For each "no-tolerance" crop/pesticide combination, our data show that FDA inspectors typically reported some of the violations, but not all. Even within the same lab, the FDA often declared some samples as illegal, yet failed to count other samples with even higher levels of the same pesticide on the same crop. For instance, acephate, an insecticide not allowed for use on tomatoes, was found on two Mexican tomatoes tested in Los Angeles in March and April of 1993. The sample with a residue at 0.008 ppm was reported as a violation, while the sample with a much higher concentration of acephate, 0.100 ppm, was not. This problem is pervasive across most FDA labs and crops tested.

Our investigation found 53 different pesticides with illegal, no-tolerance residues on just 42 fruits and vegetables. The FDA reported some of the violations for 43 of these pesticides, but the agency failed to report any illegal residues at all for 10 of the illegal chemicals.

Over-tolerance violations

The second type of illegal pesticide residue is an "over-tolerance" violation. These are pesticides that are allowed for use on the crop, but were found at levels exceeding the legal maximum level (the tolerance). The FDA did a far better job reporting the over-tolerance violations than no-tolerance violations. Even so, the discrepancy between over-tolerance residues identified by the FDA and those reported as violations, is significant.

The FDA reported just 94 over-tolerance violations, yet identified 127 over-tolerance samples, 35 percent more than it reported.

The number of over-tolerance violations is kept relatively low by the fact that tolerances for many heavily used pesticides were set in the 1950s and 1960s and have never been revised to reflect new toxicity information for the pesticide. A good example of such a pesticide is captan, which was first registered and granted tolerances in the 1950s, based on what by contemporary standards is very flimsy health and safety information. Today it is classified as a probable human carcinogen, yet the tolerances set four decades ago are still in force and run as high as 100 parts per million (ppm) for some crops, such as cherries, lettuce, plums, and spinach. To meet the EPA's one-in-one-million cancer risk standard, the tolerances for these crops would need to be reduced approximately 100-fold. If captan tolerances were lowered to meet a health-based standard, over-tolerance violations would skyrocket.

It is well established that tolerances for many pesticides exceed federal health standards (Fisher 1992a, Fisher 1992b). Examples include captan and the EBDC fungicides, where exposure at the legal limit (tolerance) exceeds federal health standards by more than 100-fold, as well as the insecticides endosulfan and methamidaphos, where exposure at the tolerance exceeds federal health standards by factors of 25 and 50 respectively. EPA allows these tolerances to remain on the book nonetheless, based on the calculation that most of the residues in the food supply are well below the tolerances and close to negligible risk levels. For some pesticides this is true, for others it is not. In total, however, combined exposures to pesticides in food far exceed negligible risk levels.

One effect of these unusually high tolerances is to desensitize FDA personnel to the occurrence of relatively high pesticide residues that are violations. Of the 15 samples with the highest illegal pesticide residues, the agency missed five. These included a U.S. onion sample with residues of captan (a pesticide canceled for use on onions) at 89 ppm (three and a half times higher than the tolerance of 25 ppm), and a peach sample with the pesticide pronamide at 8 ppm (80 times higher than the tolerance of 0.1 ppm).

The FDA, however, did report several other egregious violations. A domestic onion sample had a residue of chlorothalonil, a probable carcinogen, at 95.4 ppm (190 times the tolerance of 0.5 ppm). A green pea sample contained residues from another class of probable carcinogens, the EBDC pesticides, at 23.1 ppm (over three times the tolerance of 7 ppm), and a leaf lettuce sample contained an estrogenic pesticide with the potency of DDT, endosulfan, at 15.6 ppm (nearly eight times the tolerance of 2 ppm).

Together, "no-tolerance" and "over-tolerance" pesticides account for 79 percent of the violations identified by the FDA, and nearly all of the violations that the FDA reported. The other 21 percent of illegal pesticides are from action level excedences, trace violations, or canceled pesticides.

Action level violations

There is essentially no difference between the Environmental Working Group and FDA figures for action level violations. These are violations from banned pesticides, such as DDT, heptachlor, dieldrin, and endrin, among others. Although these pesticides were banned long ago in the United States, they degrade slowly in soil and are routinely detected in crops that grow in the soil or close to the ground. In some countries, they are still allowed for use. These pesticides do not have official food tolerances; instead they are deemed "unavoidable pesticide residues," and the FDA is given the authority to set levels below which it will not take enforcement action.

Trace violations

A trace violation is a "no-tolerance" pesticide that is detected and identified on that food at a level below which the exact amount of the pesticide can be quantified by standard FDA methods. The FDA data show 135 samples with trace violations; the agency acted on none of these.

As a matter of policy, trace violations are ignored because the agency fears that they would not hold up in court if challenged (Wessel 1995). This logic stretches credulity given the fact that illegal pesticide residues are almost never prosecuted.

Trace violations are illegal because the pesticide has been positively identified as present on the food, and because the legal level for the pesticide is zero. Nearly all trace violations could be confirmed, quantified, and acted on using off-the-shelf technologies.

Ironically, we did find five samples on which trace illegal residues were marked by the FDA as violations.

Canceled pesticides

This category of violations comes from pesticides that are canceled for use on particular foods, but where tolerances for those pesticides still remain in effect. In this situation, the FDA's hands are tied. As long as the tolerances for canceled pesticides remain on the books, foods containing these pesticides are considered legal. The FDA could, however, inform state enforcement officials of these illegal pesticide uses. To our knowledge, FDA personnel never initiated any such action.

Technically speaking, residues of canceled pesticides with tolerances still on the books are legal, even though the use has been banned in the United States. This situation allows foreign growers to use pesticides that domestic growers cannot. There is no scientific or health and safety rationale to this policy; it is simply an artifact of the way EPA (which sets and revokes tolerances) does business.

When the EPA cancels all or some uses of a pesticide it is almost always based on the determination that the pesticide should not be present on the food we eat. A December 1994 General Accounting Office report found, however, that when EPA cancels a pesticide it takes on average 6.6 years to revoke its associated tolerances. Part of this time (perhaps two years) is to allow for the depletion of existing stocks. Most of the delay, however, is due to the low priority that the EPA assigns tolerance revocation, and the lack of a policy or written guidelines to link revocation with cancellation (GAO 1994c).

During the late 1980s, the EPA canceled many of the uses of folpet and captan (a probable carcinogen and one of the most widely used pesticides). Most of the associated tolerances for these two pesticides, however, remain in the Code of Federal Regulations. The FDA identified 38 samples that contained residues of these pesticides on crops for which they are canceled in the United States. Eighteen captan violations came from domestically-grown produce, and thus represent an illegal use by the individual grower. In no situation did the FDA help initiate enforcement action against these growers. In every case the resulting food was sold to consumers.

Multiple Violations on Single Samples

No-tolerance violations accounted for 64 percent of the total violations that the FDA identified, over-tolerance violations accounted for 15 percent, action level violations totaled 1 percent, and canceled and trace no-tolerance violations made up the remaining 20 percent (Figure 2).

Some samples, however, contained more than one kind of illegal pesticide. For instance, a lettuce sample could contain a no-tolerance pesticide, chlorothalonil, and a legal pesticide, endosulfan, at a level over the tolerance. Or a sample could contain an action level violation and a no-tolerance violation. We did not include multiple violations of single samples in our overall tally of illegal samples. In calculating the 5.6 percent violation rate, all multiple violations were counted only as one illegal sample.

Multiple illegal pesticides on a single sample were not uncommon. Of the 826 samples contaminated with illegal pesticides, 82 had two illegal pesticides on the same sample, and seven had three violations. FDA detected two or more illegal pesticides on more than ten samples of strawberries, hot peppers, and green peas.


The FDA identified illegal pesticides on 5.6 percent of the 42 commonly consumed crops studied in this report. Some crops, however, were considerably more contaminated than others.

Twelve foods had violation rates greater than the 5.6 percent average (Figure 3).

Green peas and pears topped the list, with violation rates of 24.7 percent and 15.7 percent respectively. For three other commonly consumed foods--apple juice (12.5 percent), blackberries (12.4 percent) and green onions (11.7 percent)--more than ten percent of the crop contained illegal residues. Other popular foods with above average rates of illegal pesticides included hot peppers (9.7 percent), green beans (7.6 percent), strawberries (7.4 percent), and carrots (6.5 percent).

The FDA failed to report at least one illegal pesticide on 35 of the 42 different crops we analyzed, and under-reported illegal pesticides for each of the 20 commodities with the highest percentage of illegal pesticides (Figure 4). The agency missed almost half of the green pea violations, as well as 18 of 20 pesticide violations on carrots. Even worse, none of the illegal pesticides on apple juice, cranberries, and oranges were reported by the FDA. The agency made mistakes the other way as well; it reported at least 23 samples as violations when tolerances existed for the residues present.

Among the foods with the best compliance rates, avocados were free of any illegal pesticides; only one pesticide was even detected on the 136 samples. Tangerines had no violations either. Potatoes and cauliflower samples contained only one violation apiece, and grapefruit had illegal residues on just two samples. Other than potatoes and perhaps grapefruit, however, these "cleaner" foods do not account for a significant percent of the average daily intake of fresh fruits and vegetables.

Particularly worrisome is the sheer number of different illegal pesticides found on each crop. Seventeen different illegal pesticides--twelve with no-tolerance, four over-tolerance, and one canceled use--were detected on green peas over two years (Figure 5). Fourteen different illegal pesticides were found on hot peppers and squash, 12 on strawberries, 11 on carrots and pears, and 10 each on cantaloupes and bell peppers. A substantial percentage of these foods were contaminated with more than one illegal pesticide.

For Some Crops One Half of the Residue Load was Illegal

For some crops, more than 50 percent of the residues detected (the total residue load found on the crop) were illegal (Figure 6). Over 64 percent of the pesticide residues found on bulb onions were illegal, as were 51.7 percent of the residues on apple juice and 50.6 percent of those on green peas. Thirty percent of the residues found on green onions were illegal, as were 28.4 percent of the pineapple detections, 26.4 percent of the pear detections, and 22.6 percent of the carrot detections. If a pesticide is present on these foods, there is a good chance that it is illegal.


Crops From Some Countries Are Particularly Contaminated

Fifteen different crops from eleven different countries comprise the 20 worst commodity/country combinations. Seven of the 20 crops with the highest percentage of illegal pesticides came from Mexico, three from the United States, and two each from Argentina, Columbia, and Guatemala (Figure 7).

Some violation rates were incredibly high. Nearly 49 percent of the pears imported from Korea were contaminated with illegal pesticides, as were 40.8 percent of green peas from Guatemala, 34.1 percent of the peas from China, and 16.7 percent of the green onions from the United States. Violation rates were high for many important sources of major fruits and vegetables, including Mexican strawberries (18.4 percent), Mexican head lettuce (15.6 percent), Guatemalan blackberries (14.6 percent), U.S. pears (14.3 percent), Argentine pears (12.9 percent), Mexican carrots (12.3 percent), Chilean kiwi (11.7 percent), Mexican leaf lettuce (11.3 percent), U.S. tomatoes (9.4 percent), and Mexican green beans (9.4 percent).

If these violation rates are representative of the overall percentage of adulterated food, as the FDA and representatives of the produce industry routinely imply when using FDA data to support their claim that the food supply is safe, then more than one billion pounds of contaminated produce line the shelves of America's grocery stores every year.

To determine just how many pounds of produce carrying illegal pesticides could be distributed to the public, we turned to the USDA production and utilization data for 1992 and 1993. Based on the contamination rates identified by the FDA, seven popular domestically-grown fruits and vegetables contribute approximately one billion pounds of contaminated produce to the food supply each year (Table 3), including such common foods as peaches, tomatoes, and pears. Based again on FDA-identified violations, the primary supplier of imported produce, Mexico, exported to the United States an additional 82 million pounds of fruits and vegetables adulterated with illegal pesticides.

Broken down by crop, this translates to about 369 million pounds of U.S. peaches, 352 million pounds of U.S. tomatoes, 100 million pounds of U.S. green onions, 92 million pounds U.S. of pears, 83 million pounds of U.S. bell peppers, and 42 million pounds of U.S. green peas that were contaminated with illegal pesticides--yet entered the marketplace--in 1992 and 1993. From Mexico, this translates to 21 million pounds of hot peppers, 18 million pounds of watermelons, 17 million pounds of squash, 14 million pounds of strawberries, and 12 million pounds of carrots marketed with illegal pesticides during this two-year period.

Violations from Many Countries are Underreported. Even when the FDA sets up a special program to monitor problem situations, significant percentages of illegal produce slip through to the market. In 1992, the FDA set up a program targeted at repeated violations by Guatemalan snowpea growers. Despite this program, the FDA missed nearly one-half of the illegal shipments. Our analysis showed that 41 percent of the Guatemalan pea samples contained illegal residues; the FDA found only 21 percent in violation.

The countries and crops that were sampled the most (U.S., Mexico, and Guatemala) were also the most underreported (Figure 8). The actual violation rates for 16 different crops from these three countries were up to nine times higher than what the FDA reported. For instance, violation rates for Mexican carrots were seven times the FDA's reported violation rate, U.S. pears four and a half times higher, and U.S. green onions more than triple. These lists are symptomatic of the entire supply of fruits and vegetables. Twenty-five country/crop combinations were three times higher than FDA's reported violation rate and 47 combinations were double the agency's rate.


Country summary

Thirteen countries had violation rates which exceeded the overall average of 5.6 percent (Figure 9). Some of these countries are major importers, such as Mexico, Guatemala, Argentina, and Columbia.

Other countries import just a few foods, and because one or two crops contain a high percentage of violations, these countries have the highest violation rates. This is the case with Korea, Japan, China, and Greece, as well as Italy, Spain, Peru, Belgium, and Trinidad. Due to illegal pesticides on pears, imports from Korea and Japan contain the highest percentage of violations of any country we analyzed. China had problems with green peas, and Greece had violations on peaches and hot peppers.

Violations are common when countries begin exporting new commodities to the United States. These nine countries are not traditional exporters, and their growers, exporters, and governments lack information about U.S. pesticide registrations and tolerances. The FDA must pay close attention to new sources of fresh produce, sampling with more frequency and educating the appropriate players about U.S. pesticide laws.

But high violation rates are not relegated to countries unfamiliar with our tolerances. Of the 13 countries from which the FDA sampled more than 100 times over two years, four exceeded the overall average: Guatemala (24.8 percent), Argentina (10.7 percent), Columbia (8.5 percent), and Mexico (7.4 percent). Most growers and exporters in these countries should be aware of U.S. pesticide laws.

Of those countries whose crops were sampled the most, all but one had a violation rate higher than what the FDA reported (Figure 10). Again Guatemala stands out, primarily due to its problem with snow peas and blackberries. The FDA reported just 13.8 of the 24.8 percent of violations on Guatemalan produce.

More problematic are the illegal pesticides that go unreported from Mexico, Canada, and the United States, which provide the bulk of the nation's staple produce. The agency took 3,442 samples of 34 Mexican crops in 1992 and 1993, but reported only 4.2 of the 7.4 percent illegal residues it identified. The FDA underreported Mexican no-tolerance violations by one-half and over-tolerance violations by one-third. The number of illegal pesticides on Canadian produce was more than three times what the FDA reported, at 5.0 percent.

And more than one-half of the illegal pesticides on domestic fruits and vegetables went unreported. The FDA-reported violation rate for U.S. produce was 1.5 percent, but the agency identified illegal pesticides on 3.1 percent of its 6,637 samples. For some major U.S. crops, the violations were well above average, including green onions at 16.7 percent, pears at 14.3 percent, tomatoes at 9.4 percent, green peas at 8.6 percent, and peaches at 6.1 percent. These violation rates were also significantly underreported. In essence, the United States' produce supply is twice as contaminated with illegal pesticides as the FDA reports.


Sixty-six different pesticides on just 42 fruits and vegetables violated U.S. food tolerances or were illegally used in 1992 and 1993. This included 53 different pesticides with no-tolerance violations, 24 different pesticides found above established limits, two banned pesticides detected above the FDA action level, and three chemicals canceled for use in the United States but detected by the FDA (Table 4).

The FDA failed to report any violations for ten pesticides with multiple no-tolerance violations. Three of these pesticides--chlorpyrifos-methyl, imazalil, and thiabendazole--were found on six or more crops. The FDA missed five additional pesticides that were over-tolerance for all of the crops they were found on.

Some pesticides were illegally used on many different crops (Figure 11). Methamidaphos, an insecticide, was illegally used on 19 different foods. Seventeen uses had no tolerance, and two (hot and sweet peppers) were over established tolerances. Chlorpyrifos was found illegally on 16 crops and omethoate was detected illegally on 15. Acephate, captan, chlorthalonil, and permethrin were found illegally on 14 different crops.

It is clear from this analysis that illegal pesticide use is widespread. Domestic and foreign growers routinely use popular pesticides on crops for which they are not registered.