One serological and six parasitological techniques for diagnosing human trypanosomiasis were compared with regard to sensitivity, specificity, positive and negative predictive values, and practicality in field application in the Luangwa Valley of Zambia. Seven (0.64%) trypanosomiasis cases were diagnosed parasitologically in a survey of 1093 people from 19 villages. The indirect fluorescent antibody test (IFAT) was more sensitive but less specific than the parasitological techniques, detecting 71% of the confirmed cases in the first round of testing. Rat inoculation, the Giemsa stained thick film and miniature anion-exchange/centrifugation (mAEC) were all more sensitive than wet blood film examination, microhaematocrit centrifugation and wet film examination of the buffy coat after microhaematocrit centrifugation. The comparison indicated that the most effective, practical combination of techniques for survey in the Luangwa Valley was IFAT screening followed by examination of seropositive patients by rat inoculation and the mAEC (or stained thick film) in parallel format. Calculation of positive and negative predictive values showed that trypanosomiasis point prevalence measured in this way would still be underestimated by approximately 60%, indicating the need to improve IFAT specificity and parasitological sensitivity. Although only one of the seven patients diagnosed in the survey presented with signs and symptoms indicating possible trypanosomiasis, no evidence of a population of "healthy carriers" was found.