The accuracy of (1) conventional wet-mount examination, (2) Papanicolaou-stained gynecologic smears, (3) a direct slide test using fluorescein-conjugated monoclonal antibodies against Trichomonas vaginalis, and (4) two different culture media for the diagnosis of trichomoniasis in a high-risk population of 600 women was compared. Use of Feinberg-Whittington or Diamond's culture medium resulted in a diagnosis of 82 and 78 cases, respectively, and the combination of two cultures identified 88 infected women. In comparison, wet-mount examination detected only 53 (60%) of the cases. Cytologic smears were interpreted as positive for T vaginalis in 49 (56%) of the 88 cases but also resulted in seven false-positive smears, and specimens from 18 women with negative cultures were interpreted as "suspicious" for trichomoniasis. Monoclonal antibody staining detected 76 (86%) of the 88 positive specimens, including 27 (77%) of the 35 cases missed by wet-mount examination. In summary, wet-mount and cytologic studies were insensitive, and cytology study was the least specific method for diagnosis of trichomoniasis. Direct immunofluorescence with monoclonal antibodies holds promise as a sensitive and specific alternative to cultures for rapid detection of T vaginalis in clinical specimens.