We examined the impact of cognitive and biomedical variables on unprotected anal intercourse between HIV-1 infected men and casual sexual partners in a Sydney-based cohort. Participants answered questionnaires examining insertive and receptive intercourse with and without ejaculation. They completed a modified optimism-scepticism scale, a sexual beliefs scale and a clinical/demographics questionnaire. CD4 count, blood and semen VL were assessed. 43 of 109 reported anal intercourse with HIV+ partners, 33 with HIV- partners and 38 with partners of unknown status. With HIV+ partners past sexually transmittable infections were associated with receptive intercourse without ejaculation (p = 0.03) and insertive intercourse without ejaculation (p = 0.06) while sexual beliefs were associated with insertive intercourse without ejaculation (p = 0.038), receptive intercourse with ejaculation (p = 0.016) and insertive intercourse with ejaculation (p = 0.077). Sexual beliefs were found to have some association with unprotected receptive intercourse without ejaculation with HIV- partners (p = 0.071). With unknown serostatus partners, treatment-optimism (p = 0.026) had association with insertive intercourse with ejaculation while optimism (p = 0.002), sexual beliefs (p = 0.039) and recent VL (p = 0.059) had associations with insertive intercourse without ejaculation. Current STI had association with receptive intercourse with ejaculation with unknown status partners (p = 0.014). We found between-group differences in variables associated with different types of unprotected anal intercourse that may guide the development of prevention strategies.