The validity of the direct agglutination test (DAT) for visceral leishmaniasis (VL) was studied with a standardized field kit on 148 clinically suspected persons and 176 healthy controls recruited between 1993 and 1994 from an endemic area in Gedaref State, Sudan. A sensitivity of 95.9% and a specificity of 99.4% were found at a 1: 8,000 cut-off titer when parasitologically confirmed cases were compared with healthy controls. While corroborating previously reported sensitivity and specificity estimates of this serodiagnostic test, this study examined the bias generated by commonly used test validation procedures. The fundamental methodologic problem in VL test validation is the absence of a reliable gold standard. Moreover, any operational guideline on DAT use has to consider the critical dependency of the predictive values of the test on VL prevalence rates. The DAT diagnostic cut-off titer depends upon many external factors, among which the prevalence of disease in the area and the case mix seem the most important.