Leptospirosis, although ubiquitous and potentially lethal, is often not diagnosed. The seroprevalence of anti-Leptospira antibodies and the utility of two rapid tests for the serodiagnosis of the disease were studied in Binh Thuan, an area in southern Vietnam with favourable conditions for Leptospira. In an initial survey, blood samples from 44 patients with undifferentiated fever and 83 healthy subjects were each examined for anti- Leptospira antibodies using three tests: an ELISA; a latex card-agglutination test (Dri Dot); and a lateral-flow assay (LeptoTek Lateral Flow). In the ELISA, samples from 35% of the healthy subjects and 40% of the febrile patients were found to have titres of anti- Leptospira IgM of at least 1:80. Only one of the 13 patients checked again, in ELISA, 3 weeks later, showed the marked increase in IgM titre that is indicative of acute leptospirosis. In the initial survey, although the positive results of the lateral-flow assay, applied to whole blood and serum, showed a good agreement with those of the ELISA (kappa = 0.743), the results of the lateral-flow assay were often indeterminate. The card-agglutination test was more specific. The overall agreement between the results of the rapid tests and those of the ELISA was generally poor. When the samples classified as 'indeterminate' in the lateral-flow assay were considered positive, the maximum kappa-value for this assay applied to whole blood was only 0.512. In conclusion, it appears that high seroprevalences of anti- Leptospira IgM and low incidences of acute leptospirosis limit the diagnostic value of the rapid tests that were investigated. The lateral-flow assay is not specific enough. The card-agglutination test is possibly better but, because of the low incidence, its sensitivity could not be evaluated adequately in the present study.