Multiple exposures and rapidly changing use patterns are obstacles for adequate recall of pesticide exposures in epidemiologic studies. We present a simple stepwise approach for prioritization of pesticides as part of the exposure assessment strategy in an ongoing case-control study on pesticides and childhood leukemia in Costa Rica. Pesticide imports between 1977 and 2000, approximately the pertinent exposure period, were surrogates for use data. In the first phase, 323 active ingredients were identified, of which 219 were eliminated based on low usage and absence or negative results in a preliminary search in three major toxicity databases. In the second phase, the remaining 104 pesticides underwent scoring for their toxicodynamic potential (TDP) with regard to carcinogenicity, mutagenicity, and teratogenicity, weighted in this order. Bioavailability was assessed when TDP was multiplied by a weight for persistence and bioaccumulation, producing the intrinsic toxic potential (ITP). ITP was multiplied by an index of quantity (QI) of pesticide used in the exposure period, resulting in a weighted toxic potential (WTP). The top 25 positions in each of the four rankings (TDP, ITP, QI, and WTP) yielded together 64 highest-priority pesticides. This prioritization process has to be complemented with a further breakdown into crop-, time-, and biocide-specific shortlists to achieve a recall tool suitable for developing countries. Different parameters for prioritization assure inclusion of all relevant pesticides with regard to toxicity and bioavailability. The method contributes to cancer epidemiology in developing countries with access to basic use data and the Internet. The method is adaptable to other health outcomes.