As part of a screening and treatment program for intestinal parasite infections offered to newly arrived Southeast Asian refugees in Canada between July 1982 and February 1983, a total of 232 sera were tested for Strongyloides infection using an enzyme-linked immunosorbent assay (immunoglobulin G). These results were compared with coprologic results and eosinophil counts. The seroprevalence was 76.6% (131 of 171) among Kampucheans, 55.6% (15 of 27) among Laotians, and 11.8% (4 of 34) among Vietnamese. A statistically significant relation (p less than 0.001) was found between Strongyloides serology and Strongyloides infection on stool examination (prevalence, 24.7%) among Kampucheans. Eosinophilia (greater than or equal to 10%) was found to be significantly associated with both infection measures. Using coprologic results as the "gold standard," the properties of the serologic test were estimated to be: sensitivity (95%), specificity (29%), positive predictive value (30%), and negative predictive value (95%). These estimates should be regarded as minimal values, as stool examination for Strongyloides infection can be an unreliable diagnostic reference. Further evaluation of the discrepancies observed between coprologic and serologic testing is required to determine the usefulness of these tests in epidemiologic studies.