Maximum residue limits: how pesticide safety standards work

Of those samples, 43.1% contained no quantifiable residues at all, while another 54.5% contained residues within the applicable legal limits. The remaining 1.2% exceeded an MRL for at least one pesticide.

That is not the same as saying that 98.8% of all food sampled across the European Union was compliant. The figure describes the scope of the coordinated samples covered by the report. That distinction may sound technical, but food-safety statistics are often made misleading by exactly this kind of compression. A percentage is only as meaningful as the population it describes, the method used to measure it, and the limit against which it was compared.

The confusion begins with the word limit. A Maximum Residue Limit is not a toxicity threshold. It is not the dose at which a pesticide becomes dangerous, nor the invisible line beyond which an apple turns hazardous in your hand. It is a regulatory ceiling, expressed in milligrams per kilogram, describing the highest concentration of a specific pesticide residue legally permitted in or on a food or feed when that food has been treated in accordance with what regulators call Good Agricultural Practice.

The MRL tells you whether the use of the pesticide appears to have followed the approved instructions. It does not, by itself, tell you whether the food is safe to eat.

Defining MRLs: Compliance Standards vs. Toxicological Thresholds

Every pesticide registered for use on food crops carries, somewhere in its regulatory paperwork, two different kinds of numbers. The first is the MRL: the compliance standard attached to a particular pesticide, crop, and use pattern. The second is a set of toxicological safety values, principally the Acceptable Daily Intake (ADI) for long-term dietary exposure and the Acute Reference Dose (ARfD) for exposure from a single meal or a short period.

The ADI and ARfD are not derived from the MRL. They answer different questions.

• The ADI estimates the amount of a substance that can be consumed daily over a lifetime without appreciable health risk, based on the available toxicological evidence.
• The ARfD addresses the possibility of harm from a high exposure over a short period, often a single meal or day.
• The MRL establishes the residue concentration that should not be exceeded when the pesticide has been used correctly under an approved agricultural practice.

Toxicological reference values generally begin with findings from laboratory and animal studies. Regulators identify a point of departure, such as a No Observed Adverse Effect Level (NOAEL), and apply uncertainty factors to account for differences between test animals and people, as well as variation among individuals. The result is a health-based reference value intended to protect the population, including people who may be more vulnerable to chemical exposure.

The MRL is then developed through a separate process. It reflects how the pesticide behaves when applied to a crop: how much is used, how many applications are permitted, how quickly residues decline, and how long the farmer must wait before harvesting. A residue concentration at or below the MRL is therefore evidence of compliance with the approved use pattern. It is not a direct measurement of toxicity.

An MRL is the line the spray records must respect. An ADI is the line the human body must respect. They are not the same line, and they were never meant to be.

This distinction matters when a monitoring report records an exceedance. If a sample is above the legal MRL, the result may indicate that the pesticide was applied too close to harvest, used at an unauthorized rate, applied to the wrong crop, or carried over from another source. It can trigger an investigation into the farm, the applicator, the supplier, the records, or the route by which the food entered the market.

An exceedance is not automatically a finding that the food has caused, or is likely to cause, harm. Health risk assessment requires additional information: the concentration measured, the amount of food consumed, the consumer’s body weight, the toxicological reference values, and whether exposure was acute or long term. Regulators may perform a separate dietary risk assessment to determine whether the result could approach the ARfD or contribute materially to the ADI.

The unit itself also needs some texture. One milligram per kilogram is approximately one part per million. That sounds abstract because it is abstract: residue analysis operates at concentrations that are difficult to imagine in a kitchen. But the smallness of the number should not be turned into a universal safety claim. The significance of a residue depends on the substance, the food, the amount eaten, the frequency of exposure, and the relevant toxicology.

The same MRL can therefore serve two regulatory functions at once. It can identify whether agricultural instructions were followed, while also contributing data to a broader assessment of dietary exposure. It cannot replace that assessment.

The Mechanics of Calculation: From Field Trials to the OECD Calculator

How are maximum residue limits set? Not from a laboratory hunch, and not from a model alone. The number is built from supervised field trials and from the proposed agricultural use of the pesticide.

When a manufacturer seeks approval for an active substance on a new crop, residue studies are conducted under realistic but controlled conditions. The pesticide is applied according to the proposed label: the dose, the number of applications, the timing, and the pre-harvest interval. The crop is then harvested and analyzed for residues. Trials are carried out across relevant growing conditions so that the data do not describe one idealized field and one unusually cooperative season.

The resulting measurements form a distribution rather than a single answer. Some samples carry very little residue. Others show higher values because of weather, crop variety, application timing, plant growth, or ordinary variation in field conditions. Regulators use the distribution, together with the proposed use pattern and dietary exposure models, to determine an appropriate MRL.

The OECD MRL Calculator is widely used as a statistical tool in this process. Its role is often misunderstood. It does not decide whether a pesticide is toxic, and it does not turn a toxicological threshold into a legal limit. It helps regulators interpret supervised-trial data and propose a value that accommodates the expected variability of compliant agricultural use.

In broad terms, the calculation is designed around the upper end of the observed residue distribution. The objective is to avoid setting a limit so low that ordinary compliant farming repeatedly appears to violate it, while also avoiding a limit so high that meaningful deviations from the approved use pattern disappear inside the permitted range. That is a statistical compromise, but not a casual one.

Several details can affect the result:

• Application rate and timing: A spray made close to harvest may leave a different residue pattern from the same spray applied earlier.
• Number of applications: Repeated applications can change both the peak residue and the way residues decline.
• Pre-harvest interval: The waiting period allows residues to dissipate before the crop reaches the market.
• Crop variety and growth conditions: A rapidly growing crop may dilute residues differently from a slower-growing one.
• Processing and preparation: Washing, peeling, cooking, drying, or turning a crop into a processed ingredient can change dietary exposure.
• Crop grouping: Regulators may use representative crops or crop groups, but the behavior of a pesticide is not identical on every plant.
• Analytical uncertainty: Laboratories must distinguish a genuine residue from background noise and account for the limits of the method used.

This is why an MRL is better understood as the visible endpoint of a chain of assumptions. Change the approved use pattern, the crop, the climate, or the way food is consumed, and the relevant calculation may change with it.

The number is also not necessarily identical everywhere. Two jurisdictions can assess similar evidence and arrive at different legal limits because they use different crop classifications, dietary models, default rules, or interpretations of the available data. International standards developed through the Codex Alimentarius system aim to make trade and food regulation more consistent, but national and regional regulators retain authority over their own legal frameworks.

An MRL is therefore not a universal property of a chemical. It is attached to a pesticide–commodity combination within a particular regulatory system.

Regulatory Divergence: How the US and EU Manage Pesticide Tolerances

Walking through the regulatory landscape of pesticide residues means learning at least two vocabularies.

In the United States, the legal ceilings are generally called tolerances. They are established by the Environmental Protection Agency under the Federal Food, Drug, and Cosmetic Act. The Food and Drug Administration enforces tolerances for most foods, while the United States Department of Agriculture has responsibility for meat, poultry, and egg products within its regulatory sphere.

In the European Union, the corresponding term is Maximum Residue Limit. The framework involves the European Commission, the European Food Safety Authority, and the member states. Enforcement is carried out through national authorities, coordinated within the EU system.

The terminology is not merely cosmetic. The two systems can differ in how they treat a pesticide–crop combination for which no specific legal limit has been established.

The EU’s 0.01 mg/kg default is often described as a cautious floor. It does not mean that every pesticide is equally hazardous at that concentration, nor that the value is a universal toxicological threshold. It is a legal default used when a specific pesticide–crop limit has not been established.

The US system does not rely on one equivalent numerical default in every case. Where no tolerance has been set, the legal consequences of detecting a residue depend on the substance, its approval status, and the provisions that apply. An unapproved pesticide or an unapproved use can create a regulatory problem even when the concentration is very low.

This is one reason an imported shipment can be acceptable under one jurisdiction’s rules and require further scrutiny under another’s. The difference may reflect a genuinely different health assessment, but it may also arise from different agricultural practices, different assumptions about food consumption, or a different legal treatment of substances without a specific tolerance.

For farmers and traders, the practical lesson is severe but simple: a pesticide approved for use in one country is not automatically acceptable for the same crop in another. The residue limit at the farm gate is only one part of the market-access question. The destination market’s rules, analytical methods, and documentation requirements matter just as much.

Interpreting Compliance Data: Insights from the 2026 EFSA Monitoring Report

The annual EFSA report on pesticide residues is one of the most extensive public accounting exercises in food safety. The report released in May 2026 reported 98.8% compliance for the EU-coordinated sample set covered by the monitoring program. It did not establish that all food samples taken across the European Union were compliant, and it should not be read as a census of every item sold or consumed in the region.

Within the reported sample set, 43.1% contained no quantifiable residues. Another 54.5% contained residues at or below the applicable legal limits. The remaining 1.2% exceeded an MRL for at least one pesticide and became relevant for follow-up, investigation, or enforcement depending on the circumstances.

The first question when reading such a result is not whether the percentage sounds reassuring or alarming. It is: What exactly was sampled?

Monitoring programs are designed, not random acts of observation. They may include nationally selected samples, EU-coordinated samples, targeted commodities, imported foods, or products selected because earlier results suggested a reason for closer attention. The composition of the sample affects what the final percentage can tell us.

The second question is: What does “detected” mean?

A laboratory may detect a residue at a very low concentration, but detection does not imply that the residue exceeded the MRL. Modern analytical methods can identify traces far below the legal ceiling. A statement that pesticides were “found” in a food sample therefore says little without the measured concentration and the applicable limit.

The third question is: What happened after an exceedance?

An MRL exceedance may lead to origin tracing, a repeat analysis, a review of the producer’s records, or action against a shipment. It may reveal misuse, drift from a neighboring field, contamination during storage or transport, historical residues in soil, or the presence of a substance not authorized for that crop. The cause cannot be inferred from the percentage alone.

Some commodities are more likely to produce complicated residue profiles than others. Leafy vegetables have exposed surfaces and may be treated relatively close to harvest. Berries have delicate, irregular surfaces that can retain material. Crops grown in soil may encounter residues that persist from earlier agricultural use. None of these facts makes an entire category unsafe; they explain why monitoring patterns are not evenly distributed across the food supply.

The report’s numbers are most useful when read alongside the limits of the measurement:

1. Compliance is a legal classification. It tells us whether a measured concentration was at or below the relevant MRL.

2. A residue finding is not a risk estimate. Risk depends on exposure, toxicology, and consumption.

3. An MRL exceedance is not automatically an acute poisoning event. It is a signal that the approved use pattern or legal standard may not have been respected.

4. A compliant result is not proof that no chemical was present. It may mean that the residue was absent, below the laboratory’s quantification level, or present below the applicable MRL.

5. The sample frame matters. A result from EU-coordinated samples should be described as such, not expanded into a claim about every food sample in the EU.

Compliance, in EFSA’s vocabulary, means the sample respected the legal ceiling. Safety, in toxicology’s vocabulary, means the sample respected the human body’s tolerance. The two numbers describe different things, and reading the report well means holding both in mind.

For consumers, the meaningful follow-up to a headline about pesticide detection is therefore: detected at what level, relative to which limit, in which food, under which regulatory system, and with what estimated dietary exposure? Those questions are less dramatic than a single percentage, but they are much closer to the science.

The Evolving Science of Safety: TFA and Cumulative Exposure

The MRL system is not a finished instrument. It is a working framework, and the science behind it continues to change.

One example is trifluoroacetic acid, or TFA. TFA can be a breakdown product of some fluorinated pesticides and other fluorinated substances. Some of those pesticides are related to the broader group of substances known as PFAS, but PFAS are not a class of pesticides. That distinction matters: describing PFAS as a pesticide class collapses a broad chemical category into one particular use, and makes the environmental story less accurate.

TFA may persist and move through environmental systems, including soil and water. Its presence raises questions that do not fit neatly into the traditional MRL model, especially when a substance is encountered as a transformation product rather than as the original active ingredient applied to a crop.

In July 2026, EFSA announced that it had lowered the safe dietary exposure threshold for TFA by a factor of 3.5 in response to new toxicological evidence. This was not simply an MRL revision. TFA is treated as a contaminant or breakdown product in this context rather than as a conventional pesticide residue with a crop-specific application pattern. The episode illustrates a wider point: when new evidence changes the understanding of a substance’s hazard, the health-based assumptions around exposure can change even if the original agricultural-use limits have not changed in parallel.

The second issue is cumulative exposure. The MRL system was historically built around one pesticide at a time: one active ingredient, one crop, one legal number. Actual diets are less tidy. A person may consume several foods in a day, each carrying residues from different pesticides. The same active ingredient may also appear through several foods or exposure pathways.

Researchers and regulators refer to this as aggregate or cumulative exposure. The terms are related but not identical. Aggregate exposure can involve several routes, such as food and drinking water. Cumulative exposure generally concerns the combined effect of substances that may share a toxicological endpoint or mechanism.

The concern is not that every combination of residues produces a dangerous “cocktail.” That conclusion would go beyond the evidence. The point is that single-substance assessments do not automatically answer every question about combined exposure. Some chemicals may act independently; others may affect the same organ or biological pathway. The relevance depends on the substances, their concentrations, their timing, and the endpoint being assessed.

Cumulative assessment is difficult for practical reasons:

• Exposure estimates must combine different foods, preparation habits, and consumption patterns.
• Residues vary by season, geography, farming practice, and processing.
• Toxicological evidence is usually stronger for individual substances than for every possible mixture.
• A common effect does not necessarily mean that chemicals interact in a way that amplifies harm.
• Regulators must decide which substances belong in the same assessment group and which endpoints are sufficiently comparable.

EFSA and other authorities have been developing methods to address these questions. The work is technically demanding because it requires joining residue monitoring, food-consumption data, chemical grouping, and toxicology without pretending that uncertainty has disappeared.

The comparison to a financial portfolio can be useful only up to a point. The logic behind asset allocation is to distribute risk so that one holding does not dominate the outcome. In toxicology, however, residues are not investments, and the body is not a market. Chemical interactions can be nonlinear, and the mere presence of several substances does not establish a combined hazard. The analogy may help explain why regulators look beyond individual measurements, but it cannot replace the biological assessment.

Why the Framework Still Matters

Step back from the acronyms and the practical picture becomes clearer.

If you grow food, the MRL is one reference point for whether the pesticide label has been followed. Application records, spray timing, storage conditions, and harvest intervals are not bureaucratic decoration; they are the evidence behind the residue result.

If you trade food across borders, knowing which regulatory regime governs the destination is essential. A US tolerance and an EU MRL may describe a similar regulatory idea, but they are not interchangeable currencies. The legal limit depends on the crop, the pesticide, the use pattern, the jurisdiction, and sometimes the form in which the food is sold.

If you eat food and worry about residues, the useful question is not simply “are pesticides present?” Analytical chemistry can find traces at concentrations that say very little on their own. The better questions are: at what level, relative to which limit, from what source, and with what estimated exposure?

The 98.8% compliance figure from the 2026 EFSA report is not a reason to wave away legitimate questions. Nor is the 1.2% of exceedances a reason to treat every detected residue as evidence of poisoned food. The figure describes the EU-coordinated sample set covered by that report. Its value lies in showing how monitoring works: residues are measured, compared with legal standards, classified, and investigated when they cross a regulatory line.

That system has weaknesses. It can be difficult to explain, it depends on the quality and design of the sampling program, and it continues to develop as regulators confront persistent breakdown products and cumulative exposure. But it is not a single number pretending to answer every question. It is a layered system of field trials, toxicological assessment, statistical calculation, laboratory analysis, and enforcement.

The MRL is the part most visible on a chart. Behind it sits a much larger structure: the agricultural practice that produced the residue, the analytical method that measured it, the dietary model that estimated exposure, and the toxicology that gives the result meaning. Read carefully, that structure tells a more useful story than the headlines suggest.

The work is ongoing, the science is shifting, and the fine print is worth reading.