Patients attending the genitourinary medicine clinic at Watford General Hospital, UK, were examined for clinical signs of genital herpes infection. Genital swabs were taken from 194 patients (126 female, 68 male) who presented with genital ulceration or symptoms which were suggestive of genital herpes infection. Swabs from these patients were tested by three methods: (i) Detection of herpes simplex virus (HSV) antigen by direct HSV enzyme immunoassay (EIA), (ii) HSV isolation in Vero cell culture and (iii) HSV polymerase chain reaction (PCR). HSV was detected in 76 patients (39%) by EIA, in 93 (48%) by isolation in cell culture, and in 115 (59%) by PCR. Isolation by cell culture has been considered as the "gold standard" for the detection of HSV in genital lesions, but in this study HSV PCR was significantly more sensitive. Comparison of the three methods was as follows: Cell culture vs. PCR: Sensitivity 93/115 (80.9%), Specificity 79/79 (100%). HSV EIA vs. PCR: Sensitivity 75/115 (65.2%), Specificity 78/79 (98.7%). HSV EIA vs. Cell culture: Sensitivity 75/93 (80.7%), Specificity 100/101 (99%). EIA was less effective in detecting HSV among recurrent than among first episode infections, in comparison to culture or HSV PCR. This is the first comparison of HSV PCR with two other routine diagnostic methods for confirming genital herpes infection in a symptomatic population. The infecting HSV type was identified by restriction digestion of 108 HSV amplicons: HSV-1:37/108 (34%), HSV-2:71/108 (66%). In this population HSV-1 causes a significant proportion of genital herpes cases, and HSV-1 genital infection was detected in significantly more first episode infections (40.3%) than among recurrent infections (22.2%).