A substance can be perfectly harmless after a single exposure and prove toxic after four weeks of repeated administration. This is the whole point of repeated-dose toxicity studies: they explore what acute toxicity tests cannot see, namely accumulation, metabolic adaptation, progressive damage to a target organ, and the emergence of effects that only manifest after a certain exposure time.
Four OECD guidelines structure this area: tests 407 and 408 for the oral route, lasting 28 and 90 days respectively, and tests 412 and 413 for the inhalation route, over the same durations. These are lengthy, costly, and regulatory-critical studies: their results determine the NOAEL, which will serve as the basis for the risk assessment and will determine the CLP classification of your substance. Choosing the wrong duration or route of exposure means losing several months and having to start over.
This overview provides you with the selection criteria, the parameters actually measured, and how the results translate into regulatory obligations.
Table of Contents
Subacute and subchronic toxicity: what exactly are we talking about?
The difference with acute toxicity
Acute toxicity measures the effects of a single or very short-term exposure. It answers a stark, binary question: what quantity kills or causes immediate severe effects? It says nothing about what happens when a worker breathes in dust every day for months, or when a consumer ingests trace amounts of a residue daily.
Repeated-dose studies answer this second question. They do not seek a lethal dose but a threshold without observed adverse effects, by administering deliberately non-lethal doses for a prolonged period, and looking for the first signs of functional or tissue damage.
Subacute or subchronic: a question of duration
The terminology is internationally standardized. Subacute to repeated exposure over approximately 28 days, and subchronic to exposure over 90 days, roughly one-tenth of a rodent's lifespan. Beyond that, we enter the realm of chronic toxicity, which is covered by other guidelines.
This difference in duration is not merely cosmetic. Certain effects—liver enzyme induction, renal tubular damage, thyroid disturbance, slow degenerative lesions—simply do not appear within 28 days. A 28-day study that detects nothing therefore does not prove the safety of a substance with prolonged use; it only proves that nothing was detectable within that timeframe.
What the study is really looking for
A repeated-dose study pursues four simultaneous objectives:
- Identify the target organs : which tissue is affected first? Liver, kidney, thyroid, hematopoietic system, respiratory system?
- Establish a dose-response relationship : does the effect increase with the dose, and by what slope?
- Determine the NOAEL and the LOAEL, which will become the starting points for calculating exposure limit values.
- To assess the reversibility of the observed effects, thanks to the so-called "satellite" groups maintained without treatment after the exposure period.
The four guidelines at a glance
| Criteria | OECD 407 | OECD 408 | OECD 412 | OECD 413 |
|---|---|---|---|---|
| Way | Oral | Oral | Inhalation | Inhalation |
| Duration | 28 days | 90 days | 28 days | 90 days (13 weeks) |
| Kind | Subacute | Subchronicle | Subacute | Subchronicle |
| Preferred species | Rat | Rat | Rat | Rat |
| Animals in groups | 5 males + 5 females | 10 males + 10 females | 5 males + 5 females | 10 males + 10 females |
| Dose groups | At least 3 levels + 1 control group (air and/or vehicle for inhalation) | |||
| Exposure | Daily | Daily | 6 hours/day, 5 days/week | 6 hours/day, 5 days/week |
| Satellite group | Optional, ≥ 14 days of recovery | Optional, ≥ 14 days | Optional, ≥ 14 days | Optional, ≥ 14 days |
| Current version | 2008 | 2018 | 2018 | 2018 |
Oral route: OECD trials 407 and 408
OECD 407 — Repeated-dose oral toxicity, 28 days
Trial 407 constitutes the entry point for the repeated-dose evaluation. The substance is administered daily by gavage, or incorporated into the feed or drinking water, for 28 days, at at least three dose levels plus a control group, to groups of ten animals divided equally between males and females.
The protocol includes daily clinical monitoring, supplemented in the fourth week by a battery of functional observations: sensory reactivity to stimuli, grasping strength, motor activity. These are complemented by hematology, clinical biochemistry, organ weighing, and then a histopathological examination covering an extensive list of tissues — brain, spinal cord, eye, stomach, intestines, liver, kidneys, adrenal glands, spleen, heart, thymus, thyroid, trachea, lungs, gonads and accessory sexual organs, bladder, lymph nodes, peripheral nerve, skeletal muscle, bones and bone marrow.
The 2008 revision significantly enhanced this guideline by incorporating parameters sensitive to endocrine function, particularly thyroid function. This is what earned this version the nickname " enhanced TG 407 " in the literature: beyond general toxicity, the test now provides warning signals about potential endocrine-disrupting activity, without, however, constituting an endocrine disruption test on its own.
OECD 408 — subchronic oral toxicity, 90 days
Trial 408 follows the same logic over 90 days, with double the sample size: ten males and ten females per group. This increase is not a mere formality—it gives the study the statistical power necessary to detect small-amplitude effects, precisely those that would escape detection in a 28-day study.
The current version, revised in 2018 and then reissued in 2025, also incorporates a series of measures related to the endocrine system, with particular attention paid to thyroid function. Trial 408 is now the reference study for establishing a usable NOAEL in a risk assessment, and it is the one that regulatory agencies expect when tonnages or uses warrant it.
We dedicate a full article to the detailed course of this trial: OECD 408: 90-day subchronic oral toxicity study in rodents.
Why the 407 never replaces the 408
This is the most frequent, and most costly, strategic error. Test 407 is often perceived as a budget version of 408. It is not.
Its small sample size and short duration result in significantly lower sensitivity. A negative result after 28 days does not allow for the conclusion that there is no effect from prolonged exposure, and the authorities will not accept it as a substitute when the 90-day study is required. However, test 407 fulfills two functions very well: it guides the choice of doses for the 90-day study, avoiding the selection of a high dose that would cause excessive mortality, and it serves as an initial warning when nothing is yet known about a substance.
Ordering a 407 followed by a 408 costs more than ordering a 408 directly. But blindly launching a 408 with poorly calibrated doses, and having to redo it, costs much more.
Respiratory route: OECD trials 412 and 413
An exhibition method that follows its own rules
Inhalation tests are not simply transpositions of oral tests. Exposure is described not by an administered dose but by an atmospheric concentration, and the fate of the substance depends closely on the size of the particles generated.
The guidelines stipulate that aerosols must have a mass median aerodynamic diameter (MMAD) between 1 and 3 µm, with a geometric standard deviation between 1.5 and 3.0. This requirement is not a mere technical detail: it ensures that the particles effectively reach the rodent's deep respiratory tract. A poorly characterized atmosphere invalidates the entire study, regardless of the biological results obtained.
OECD 412 — subacute inhalation toxicity, 28 days
The animals are exposed for six hours a day, five days a week, for four weeks, to at least three concentrations plus a control exposed to air or vehicle. The main study uses five males and five females per group.
Beyond the standard parameters—clinical observations, body weight, food and fluid intake, hematology, biochemistry, organ weighing, and histopathology—the test may include bronchoalveolar lavagewith total and differential white blood cell counts, total protein levels, and lactate dehydrogenase levels. These markers detect early pulmonary inflammation, even before a histological lesion is visible.
OECD 413 — subchronic inhalation toxicity, 90 days
The same exposure schedule—six hours a day, five days a week—was maintained, but spread over thirteen weeks, with ten males and ten females per group. A satellite group of the same size, exposed to the highest concentration, allowed for the assessment of reversibility after at least fourteen days without exposure.
The protocol includes an ophthalmological examination before and after the trial. Bronchoalveolar lavage is recommended when the lower respiratory tract is the primary site of deposition. Additional investigations may be incorporated depending on the specific issues: kinetics, biomonitoring, pulmonary function testing, retention of poorly soluble materials, and behavioral observations.
From result to classification: the link with CLP
This is the point that many manufacturers discover too late. The NOAEL from the study is not just a figure in a report: it directly determines whether your substance will need to carry a STOT-RE (specific target organ toxicity, repeated exposure).
The CLP regulation sets guideline values, expressed for a 90-day study in rats. A significant toxic effect observed at a dose less than or equal to these thresholds leads to a classification.
| Exhibition route | Category 1 | Category 2 |
|---|---|---|
| Oral (rat) | C ≤ 10 mg/kg bw/day | 10 < C ≤ 100 mg/kg bw/day |
| Cutaneous (rat/rabbit) | C ≤ 20 mg/kg bw/day | 20 < C ≤ 200 mg/kg bw/day |
| Inhalation — gas | C ≤ 50 ppmV/6 h/day | 50 < C ≤ 250 ppmV/6 h/day |
| Inhalation — vapors | C ≤ 0.2 mg/L/6 h/day | 0.2 < C ≤ 1.0 mg/L/6 h/day |
| Inhalation — dust, mists, fumes | C ≤ 0.02 mg/L/6 h/day | 0.02 < C ≤ 0.2 mg/L/6 h/day |
An extrapolation rule should be noted: for a 28-day study, these guideline values are multiplied by three. An effect observed at 25 mg/kg/day in a 407 trial thus falls within the category 1 range (threshold of 30 mg/kg/day at 28 days), whereas the same effect at 25 mg/kg/day in a 408 trial would fall within category 2. The duration of the study therefore changes the regulatory outcome, even with an identical biological effect.
These values are guidelines, not automatic thresholds: the final classification results from an expert judgment integrating the nature and severity of the effects, their relevance to humans and the consistency of all the data.
When are these tests required?
Under REACH, a tonnage-based approach is used
The REACH regulation links information requirements to the volume placed on the market. In practice, a 28-day becomes mandatory for quantities of 10 tonnes per year or more per registrant, while a 90-day subchronic is required once the threshold of 100 tonnes per year is exceeded. For quantities exceeding 1,000 tonnes, longer-term studies may be required.
The chosen exposure route must reflect actual human exposure. The oral route is the default choice, but if workers or consumers are primarily exposed via inhalation—powders, aerosols, volatile substances—then tests 412 or 413 are required, not a simpler oral study.
Beyond REACH
These same guidelines structure the dossiers for biocidal products, plant protection products, food and feed additives, as well as the biological evaluation of medical devices. The cosmetics sector is an exception: since animal testing is prohibited in this sector under Regulation 1223/2009, such tests cannot be carried out for cosmetic safety assessment purposes within the European Union.
Are you looking for an analysis?
How to choose: the four questions to decide
By what route does human exposure actually occur? This is the primary question, and it determines the pair 407/408 or 412/413. An answer based on experimental convenience rather than on the actual exposure scenario weakens the entire case.
What is my tonnage today and in three years? A company approaching the 100-tonne threshold would be better off directly considering the 90-day option rather than financing a 28-day option that will need to be supplemented shortly afterwards.
What do we already know about the substance? In the complete absence of repeated dose data, a 28-day study or a preliminary dose-finding study secures the choice of concentrations for the main study.
Is reversibility an issue? If an effect is expected, including a satellite group from the design stage makes it possible to demonstrate that the impairment regresses when exposure is stopped — an argument that is often decisive in discussions with the evaluator.
Three key areas of concern that can cause a case to fail
A poorly calibrated dose selection. A high dose that causes mortality or excessive weight loss renders the group uninterpretable; a high dose that is too low produces no effect and does not allow for characterizing the dose-response relationship. In both cases, the study fails to fulfill its purpose.
An insufficiently characterized exposure atmosphere. For tests 412 and 413, the particle size distribution, homogeneity, and stability of the atmosphere must be continuously documented. This is the first thing an assessor checks, and the primary cause of rejection.
Histopathology not compliant with GLP. These studies must be conducted according to Good Laboratory Practices to be accepted in a regulatory dossier. A histopathological reading performed outside the scope of GLP, or by an unauthorized laboratory, invalidates several months of animal work.
Frequently Asked Questions
Can a 90-day study be replaced by a 28-day study?
No, not when the 90-day period is legally required. The 407 and 412 standards have insufficient statistical power and observation window to detect slowly evolving effects. They guide dose selection and serve as an initial warning, but are not an acceptable substitute.
How long should a 90-day study be scheduled for?
The exposure phase lasts thirteen weeks, but this must be supplemented by the preparatory phase—analytical characterization, preliminary dose studies, and setting up the inhalation atmosphere—followed by histopathology and the preparation of the GLP report. An overall timeframe of six to twelve months is realistic, depending on the complexity of the substance and the availability of the laboratory.
Are there any non-animal alternatives to these tests?
No validated method currently exists that can completely replace a repeated-dose study. However, new methodological approaches—in vitro organ models, toxicokinetic modeling, read-across approaches, and chemical categories—make it possible to reduce the number of studies required, guide experimental design, and, in some cases, justify adapting reporting requirements. These strategies must be carefully documented to be accepted.
Why are these tests now measuring thyroid parameters?
Because successive revisions of guidelines 407, 408, 412, and 413 have incorporated measures sensitive to endocrine function, with thyroid function being of particular concern. The aim is to detect an early warning signal within a general toxicity study, which can then trigger further specific investigations. These parameters do not, however, transform the trial into a full-fledged endocrine disruption test.
What species is used?
The rat is the preferred species for all four guidelines. Using another species remains possible but must be scientifically justified, and it complicates comparison with available historical data.
Should a satellite group always be included?
It is optional, but strongly recommended whenever an effect is expected at the high dose. A recovery period of at least fourteen days allows for differentiation between a transient effect and a persistent lesion—a distinction that carries significant weight in the final assessment of the risk.
Conduct your repeated-dose toxicity studies with YesWeLab
These trials involve significant budgets, lengthy timelines, and animal experimentation that no one wants to repeat. The quality of the laboratory, its GLP status, and its ability to accurately characterize exposure are not variables that can be adjusted.
YesWeLab connects you with accredited partner laboratories, selected for their mastery of OECD guidelines and their compliance with Good Laboratory Practices. We support you from the outset on the most crucial aspect: choosing the exposure route, study duration, and dose plan tailored to your substance and regulatory requirements. Our platform centralizes your requests, quotes, and reports, and we remain your single point of contact throughout the entire study.
You can view all of our services in our analysis catalogue, or explore the other OECD guidelines we cover.
Make YesWeLab your go-to choice for your analyses and focus on what you do best.
