OBJECTIVE: To evaluate the extent to which 'high-risk' sexual behaviour is influenced by awareness of partners' HIV status among gay men. DESIGN: Structured interviews and collection of saliva samples for anonymous linked testing for HIV-1 antibodies. SETTING: Genitourinary medicine clinics and the gay community. SUBJECTS: Men (n = 677) who reported sexual contact with another man in the last 5 years. RESULTS: The majority of respondents (63%) had had an HIV-antibody test. Analysis of data showed that in 15% of the respondents' 1380 partnerships, HIV status was known by both parties. However, the majority of partnerships involved only safe sex. Only 26% of the partnerships in which unprotected penetrative anal sex had occurred involved mutual knowledge of HIV status and was most likely to occur with regular rather than non-regular/causal partners. Logistic regression revealed that this latter association could not be explained in terms of mutual HIV status knowledge. CONCLUSIONS: Despite widespread HIV testing, the majority of gay men engaging in high-risk sex are unaware of their partner's HIV status.

PIP: The authors interviewed and collected saliva samples from 677 men who reported having sexual contact with other men in the preceding five years to assess the extent to which high-risk sexual behavior is influenced by awareness of partners' HIV status among gay men. Participants of mean age 32.6 years of a range of 16-71 years were recruited from genitourinary medicine clinics and the gay community over the period March 1991-April 1992. 52% were from London, 21% from Manchester, 18% from central England, 6% from Bristol, and 3% from elsewhere. 63% had previously had an HIV-antibody test. 94 (16%) of the 580 satisfactory saliva samples tested for HIV-1 antibodies were seropositive; 17 men were previously unaware of their HIV-positive serostatus. The 577 men who reported having a sex partner within the previous month provided information on 1380 partners during the period. 45% of the sexually active men reported engaging in some kind of penetrative anal sex in the previous month, while 23% reported having unprotected anal sex in the previous month. HIV status was known by both parties, however, in only 15% of the 1380 partnerships. 891 partnerships involved no penetrative anal sex and 1107 involved only nonpenetrative or penetrative sex always with a condom. Further, 26% of the partnerships in which unprotected penetrative anal sex took place involved mutual knowledge of HIV status and was most likely to occur with regular instead of nonregular/casual partners. It is of interest that 114 partners were known to be married and 311 were of unknown marital status. Higher social class was associated with a reduced likelihood of having risky sex and high-risk sex was more common in partnerships in which there was an age gap in excess of two years between partners.