Objectives Post-exposure prophylaxis (PEP) care remains a challenge for individuals with potential sexual exposure to HIV in terms of PEP completion and ongoing risk behaviours. Methods A retrospective analysis was carried out on data from the French Dat’AIDS prevention cohort (NCT03795376) for individuals evaluated for PEP between 2004 and 2017. A multivariable analysis was performed of predictors of both PEP completion and condom use [odds ratios (ORs)] and their associated probabilities (P, with P > 95% being clinically relevant). Results Overall, 29 060 sexual exposures to HIV were evaluated for PEP [36% in men who have sex with men (MSM) and 64% in heterosexuals]. Overall, 12 different PEP regimens were offered in 19 240 cases (46%). Tenofovir disoproxil fumarate (TDF)/emtricitabine (FTC) was the preferred backbone (n = 14 304; 74%). We observed a shift from boosted protease inhibitor-based regimens to nonnucleoside reverse transcriptase inhibitor- or integrase inhibitor-based regimens in recent years. Overall, 20% of PEP prescriptions were prematurely discontinued. Older age, MSM, intercourse with a sex worker, rape and intercourse with a known HIV-infected source patient were factors associated with increased rates of PEP completion (OR > 1; P > 98%). None of the 12 PEP regimens was associated with premature discontinuation. We also found 12 774 cases of unprotected sexual intercourse (48%). Condom use decreased (OR < 1; P > 99%) with the year of exposure, and was lower in MSM and rape victims. Condom use increased (OR > 1, P > 99%) with age, and was higher in those who had intercourse with a sex worker or with a female partner and in those with knowledge of the partner’s HIV status. Conclusions We provide new insights into how rates of condom use and PEP completion might be improved in those receiving PEP by targeting certain groups of individuals for interventions. In particular, youth and MSM at risk should be linked in a prevention-to-care continuum.