An official website of the United States government.

We've made some changes to EPA.gov. If the information you are looking for is not here, you may be able to find it on the EPA Web Archive or the January 19, 2017 Web Snapshot.

Pesticide Registration

Guidance on Efficacy Testing for Pesticides Targeting Certain Invertebrate Pests

The information provided below is intended to help pesticide registrants or applicants for registration complete the efficacy requirements related to pesticides intended to combat invertebrate pests. This guidance is not binding on either EPA or any outside party, and the EPA may depart from the guidance where circumstances warrant without prior notice. Registrants may propose alternative approaches to the recommendations in this guidance, and the Agency will assess them for appropriateness on a case-by-case basis.

On this page:
  • Efficacy label claims and associated efficacy testing
    • What efficacy data should be provided to support a “knockdown,” “quick kill” or “kills on contact” claim?
    • What species should be tested to obtain approval for label claims against ticks or mosquitoes?
    • What efficacy data should be submitted to obtain approval for swimming, shampooing or waterproof claims for pet spot on products?
    • What substrates should be tested to support residual efficacy label claims?
    • What life stage should be tested to support efficacy claims for a pest if no life stage is specified on the label?
  • Experimental design for efficacy studies
    • For dog and cat on-animal studies, when should flea and tick counts be taken?
    • Should dead and moribund arthropods be combined to obtain a percent mortality calculation?
    • For testing the efficacy of direct sprays and residual deposits in the laboratory, how many replicates are recommended?
    • Should geometric or arithmetic means be used when reporting efficacy results?
    • Should inerts-only controls be tested in addition to untreated controls for pet spot-on efficacy studies?
    • Should test specimens be transferred to clean containers after pesticide treatment?
  • Efficacy data and protocol submission
    • Under what circumstances should an efficacy protocol be submitted, and what information should be included in the draft protocol?
    • What raw data should be submitted with efficacy studies?
  • For more information

Efficacy Label Claims and Associated Efficacy Testing

What efficacy data should be provided to support a “knockdown,” “quick kill” or “kills on contact” claim?

To make “knockdown,” “quick kill” or “kills on contact” claims, data should be provided that show: 

  • ≥90% knockdown within 10 seconds for stinging Hymenoptera (including fire ants) or within 30 seconds for all other arthropods; and
  • ≥90% mortality by 96 hours post-treatment. 

In addition, similar claims may be made with a time qualification (e.g., “kills quickly within 10 minutes”) if data show:

  • ≥90% knockdown at the labeled time qualification; and
  • ≥90% mortality by 96 hours post-treatment.

Under both scenarios, treatment groups should be statistically different from control groups, and control mortality should remain ≤10%.

What species should be tested to obtain approval for label claims against ticks or mosquitoes?

To obtain approval for any labeling claims against ticks:

  • for products other than dog or cat products, efficacy data should be submitted for the following three ticks:
    • blacklegged tick (Ixodes scapularis);
    • lone star tick (Amblyomma americanum); and
    • either American dog tick (Dermacentor variabilis) or brown dog tick (Rhipicephalus sanguineus).
  • on dog products, efficacy data should be submitted for the following three ticks:
    • blacklegged tick (Ixodes scapularis);
    • the American dog tick (Dermacentor variabilis); and
    • the brown dog tick (Rhipicephalus sanguineus).
  • on cat products, efficacy data should be submitted for the following three ticks:
    • blacklegged tick (Ixodes scapularis);
    • the American dog tick (Dermacentor variabilis); and
    • the lone star tick (Amblyomma americanum).

To obtain approval for any labeling claims against mosquitoes, efficacy data should be submitted for the following three genera:

  • Anopheles (Anopheles quadrimaculatus or Anopheles freeborni or Anopheles punctipennis or Anopheles gambiae); 
  • Aedes (Aedes albopictus or Aedes aegypti); and 
  • Culex  (Culex pipiens or Culex quinquefasciatus or Culex tarsalis). 

Testing of the specified three tick species/three mosquito genera ensures that there are data supporting efficacy against the major disease vectors in these groups (e.g., Ixodes scapularis for Lyme disease, Aedes albopictus/Aedes aegypti for Zika virus, etc.). Requiring data on major vectors is necessary to ensure that pesticide products are effective against species that may pose risks to public health.

What efficacy data should be submitted to obtain approval for swimming, shampooing or waterproof claims for pet spot-on products?

An acceptable shampooing study is one that uses a non-medicated shampoo on the companion animal followed by rinsing of the animal for at least 5 minutes using a minimum of a 3.0 gallon per minute rain showerhead. The resulting data should demonstrate at least 90% efficacy against the target species.

Swimming and waterproof claims would also be allowed using this type of study. For swimming or waterproof claims only, a protocol should be submitted prior to testing.  

What substrates should be tested to support residual efficacy label claims?

The choice of substrate is dependent on the labeled sites and claims:

  • If a product is intended for use as an indoor residual treatment, both a porous (unpainted/unfinished ¼" thick plywood) and a nonporous (glazed ceramic tile) should be used.
    • If fabric, mattresses, or carpet are included as a labeled site, efficacy data using cotton sheeting, mattress ticking and/or carpeting should also be provided.
  • If a product is intended for use on buildings or as an outdoor residual perimeter or exterior crack and crevice treatment, a non-porous surface (vinyl siding or tile) and a porous surface (unpainted/unfinished ¼" thick plywood or unpainted concrete) should be used in product performance testing; if relevant to the labeled use patterns, concrete should be used instead of plywood. 
  • If a product is intended for use only on pavement, a porous surface (unpainted concrete) should be used. 
  • If a product is intended for use on trees, shrubs, hedges, etc., recently collected leaves should be included as a test surface. The plant species tested should be easily propagated and common in the areas where the product will be used. Leaves should be clipped from the treated plant as close to the exposure period as practical.

Treated surfaces should be stored outdoors such that they are exposed to direct sunlight and precipitation. Indoor aging regimes with simulated outdoor conditions should include diel cycles of photoperiod with a light source containing both UVA and UVB and periodic simulated rain, in addition to considering appropriate temperature and relative humidity. Regardless of whether the aging takes place outdoors or is simulated in the laboratory, the conditions used should mimic those under which the product would typically be used.

Please see the revised Product Performance Test Guideline OSCPP 810.3500 Premises Treatment.

What life stage should be tested to support efficacy claims for a pest if no life stage is specified on the label?

If a proposed labeling claim specifies a pest but not a particular life stage (e.g., "kills fleas"), adults should be used to obtain the claim.

Top of Page

Experimental Design for Efficacy Studies

What application rate should be tested in spot-on efficacy studies?

The lowest labeled dose should be used in efficacy studies supporting spot-on products. The lowest labeled dose refers to the lowest pesticide application rate to be applied to any pet per the label instructions. If a product is labeled for dosing by weight class (e.g., apply 0.34 fl. oz. tube of product to dogs weighing between 7 and 14 lb), the lowest labeled dose is the label dose (0.34 fl. oz.) divided by the maximum weight within the weight class (14 lb). Per the example above, the lowest labeled dose is 0.34 fl. oz./14 lb dog or 0.0243 fl. oz./lb. If a product employs dose banding, wherein the product allows for variable dosing determined by the weight class of the pet, the lowest labeled dose should be calculated for each weight class. The lowest labeled dose for the product is the lowest dose across all weight classes. By using the lowest labeled dose in efficacy studies, data will be generated to support all animals within a weight class.

For dog and cat on-animal studies, when should flea and tick counts be taken?

Flea and tick counts should be taken at 24 and/or 48 hours:

  • post-treatment for the initial infestation; and
  • post-infestation for subsequent reinfestations.

If 72-hour counts are justified based on the active ingredient mode of action, this departure and justification should be described in the submitted study.

Should dead and moribund arthropods be combined to obtain a percent mortality calculation?

No. EPA does not consider knocked down or moribund arthropods to be dead (i.e., motionless even when probed).

The number of dead, knocked down, and live pests in each replicate should be recorded separately at each time point tested, as practically possible. At a minimum, for mortality calculation purposes, dead individuals should be recorded at the last time point tested or at 96 hours after the onset of application/initial exposure, whichever comes first. A mortality count should include only dead, but not knocked down, arthropods.

While moribundity is not used as an endpoint for efficacy studies, if moribund individuals in a study are to be counted and reported, a definition of moribund appropriate for the species tested should be provided in the study report; moribund individuals should not be considered dead or be part of a mortality count.

For testing the efficacy of direct sprays and residual deposits in the laboratory, how many replicates are recommended?

Unless otherwise justified by the results of a power vs. sample size analysis, balanced (equal number of treatment and control replicates) experimental designs using a minimum of five replicates of 15 individuals, seven replicates of 10 individuals, or 35 replicates of one individual for each treatment, species, and, if applicable, surface type are recommended for most studies. Due to potential pest availability challenges, 25 replicates of one individual may be an appropriate replication scheme for centipedes, spiders, and scorpions, although 35 replicates of one individual is preferred. The protocol should fully describe how sample size and replication were determined. Pseudo-replication should be avoided; for example, an experiment in which five petri dishes of insects are sprayed using one sweep of a single aerosol can should not be considered to have five replicates. For exceptions and additional information, please see the revised Product Performance Test Guideline OSCPP 810.3500 Premises Treatment.

Should geometric or arithmetic means be used when reporting efficacy results?

The Agency prefers the use of arithmetic means for efficacy studies. For many studies, counts are the true response variable (percent control is considered a transformed response variable), and count data are often discrete and follow a Poisson (or binomial) distribution, which are more accurately summarized using the arithmetic mean. This is especially true for studies where a set number of individual insects are tested. 

Geometric means are more appropriate for continuous data that follow a log-normal distribution. If the geometric or other mean will be used to summarize the data, then evidence that the data follow an appropriate distribution should be provided to show that the statistic is valid.

Should inerts-only controls be tested in addition to untreated controls for pet spot-on efficacy studies?

For most studies, an inerts-only control would not be needed as a negative control.

Should test specimens be transferred to clean containers after pesticide treatment?

Yes. Test specimens should be transferred to clean containers after pesticide treatment because continuous exposure to a pesticide treatment is not a realistic scenario in most cases (one exception is treatment of immobile life stages (e.g., eggs)):

  • For direct application studies (i.e., the specimen is treated directly with a pesticide), specimens should be transferred to clean containers as soon as practical but no more than 15 minutes after onset of exposure to pesticide application.
  • For residual application studies (i.e., a surface is treated with a pesticide and the specimen is subsequently exposed to the treated surface), length of pest exposure to treated surfaces should be minimized and mimic real-world exposures based on the species tested. Crawling pests should be exposed to treated panels for no more than four hours. Flying pests should be exposed to treated panels for no more than one hour.

Please see the revised Product Performance Test Guideline OSCPP 810.3500 Premises Treatment for exceptions and more information.

What application rate should be tested in spot on efficacy studies?

The lowest labeled dose should be used in efficacy studies supporting spot on products. The lowest labeled dose refers to the lowest pesticide application rate to be applied to any pet per the label instructions. If a product is labeled for dosing by weight class (e.g., apply 0.34 fl. oz. tube of product to dogs weighing between 7 and 14 lb), the lowest labeled dose is the label dose (o.34 fl. oz.) divided by the maximum weight within the weight class (14 lb). Per the example above, the lowest labeled dose is 0.34 fl.oz./14 lb dog, or 0.0243 fl. oz./lb. If a product employs dose banding, wherein the product allows for variable dosing determined by the weight class of the pet, the lowest labeled dose should be calculated for each weight class. The lowest labeled dose for the product is the lowest dose across all of the weight classes. By using the lowest labeled dose in efficacy studies, data will be generated to support all animals within a weight class.

Top of Page

Efficacy Data and Protocol Submission

Under what circumstances should an efficacy protocol be submitted and what information should be included in the draft protocol?

For non-guideline or other innovative experimental designs, novel products, or uncommon labeling claims, protocol submissions are helpful to assure registrants that their proposed methods are scientifically sound and able to produce reliable data to support their desired claims. Review of such protocols also helps ensure that test results would conform to current scientific standards.

Registrants should submit detailed descriptions of their objectives, experimental design for meeting such objectives, and discussion of appropriate data analysis so that a reviewer could perform the experiment by following the methods described in the submitted protocol. In addition, proposed labeling claims or a draft label should be submitted with the protocol.

For research involving intentional exposure of human subjects, study protocols must be submitted for joint review by EPA and the Human Studies Review Board (HSRB) and be approved prior to performing the studies in accordance with EPA’s regulations (40 CFR part 26).

What raw data should be submitted with efficacy studies?

40 CFR §160.3 describes the Agency’s definition of raw data. Generally, to review an efficacy study, efficacy reviewers should have:

  • count data for treated and control replicates (including data sheets); and
  • statistical analyses.

Top of Page

For More Information

If you have questions, send us an email (OPP_RD_INVERT_EFFICACY@epa.gov).

Top of Page