Variations in estimates of prevalence of trichomoniasis in men may reflect true differences in the burden of disease but are also affected by the performance of diagnostic methods and the type of specimen tested. In this study, men were evaluated at baseline and at follow-up, to evaluate syndromic management of urethritis and the effects of human immunodeficiency virus and trichomoniasis, in Lilongwe, Malawi. First-void urine specimens and urethral swabs were obtained at enrollment, for Trichomonas vaginalis culture; semen specimens were also obtained at follow-up. The sensitivities of testing methods using urine specimens and urethral swabs were equal; 67% of cases were identified by use of either specimen, and, in 47% of cases, both specimens tested positive. When semen specimens were included, all 3 specimens tested positive in only 19% of cases. Semen was the most sensitive single specimen, and, in 25.6% of cases, only semen specimens tested positive. Thus, prevalence of T. vaginalis infection in men is underestimated if only 1 specimen is tested.