Seven laboratories participated in a collaborative study to evaluate a proposed EPA standard protocol for the Ames test. The study utilized Salmonella typhimurium strains TA98 and TA100 with three metabolic activation levels (0, 2, and coded unknowns. The three primary study goals were to evaluate the EPA protocol, to determine if the protocol produced qualitative results that agreed with published qualitative results, and to determine the qualitative and quantitative reproducibility of results. Regarding the first goal, most of the participants agreed that the EPA protocol was technically acceptable, although labor-intensive. The second and third goals of the study are interrelated. Ability to obtain qualitative results in agreement with published data was less than that reported in an earlier study by de Serres and Ashby (1981) in which each laboratory used its own protocol. The conclusion from analysis of the quantitative data from the interlaboratory Ames study was that both intralaboratory and interlaboratory variation were substantial.