The applicability of a circulating Schistosoma antigen detection assay for determining rates of infection and efficacy of chemotherapy was evaluated in Mali. Urine egg counts were compared to circulating anodic antigen enzyme-linked immunosorbent assay (CAA-ELISA) titres in serum, before and 6 weeks after treatment with a single dose of praziquantel (40 mg/kg), in 2 villages in Dogon Country, an area endemic for S. haematobium, the predominant schistosome infection in Mali. In Kassa, a village with a moderate prevalence of infection, the serological prevalence (48%) was significantly higher than the parasitological prevalence (31%). In Boro, a village with high parasitological prevalence (76%), no difference was observed between the results of both methods (prevalence by CAA-ELISA was 75%). Cure rates estimated by CAA-ELISA were lower than those determined parasitologically, suggesting that cure rates are overestimated by egg counting. The sensitivity of the CAA-ELISA was 78%. In both villages, before treatment, a positive correlation was found between the number of eggs in urine and serum CAA titres. It is concluded that, although further simplification and improvement of the sensitivity of the assay is needed, in its present ELISA format the antigen detection assay is useful for monitoring sentinel populations. Furthermore, the serum CAA assay performed adequately in a public health laboratory within an endemic country.