1. Availability and possession of condoms, condom negotiation, and planning to use condoms were event-specific factors related to lower likelihoods of unprotected anal intercourse (UAI) with both regular and non-regular male sex partners (RP and NRP). Situational risk factors for UAI were anal sex taking place during the weekday and tiredness prior to anal sex. Other situational and environmental factors were display of reminders promoting condom use, home settings, travel, and nervous feelings.
2. HIV prevention should consider the aforementioned factors and target specifically on inconsistent condom users. It should therefore be based on research discriminating between episodes of UAI and protected anal intercourse (PAI) within inconsistent condom users, rather than those obtained from traditional studies discriminating between consistent and inconsistent condom users. This is especially important as our findings suggested that the factors derived by the two types of studies might differ.
3. HIV prevention should consider segmentation of RP and NRP, as predictors of UAI differ among the two partner types.
4. This study demonstrates the usefulness of a case-crossover design for investigating event-specific variations in behaviours.