Low perception of HIV risk is a challenge to PrEP implementation. We analyzed associations between perceptions of PrEP candidacy, behavioral indications for PrEP, and sexual behaviors. We recruited a sample of 580 MSM from a geosocial-networking smartphone application in Paris, France. A modified Poisson regression model was conducted to examine associations between perceived candidacy for PrEP and behavioral indications for PrEP, and relationships among engagement in group sex, transactional sex, HIV test history, and indications for PrEP. Adjusted risk ratios (aRR) and 95% confidence intervals (CIs) were calculated. For the outcome of perceived candidacy for PrEP, a multinomial logistic regression was performed, and adjusted relative risk ratios (aRRR) were calculated. Multivariate analyses were adjusted for socio-demographics. Respondents who considered themselves PrEP candidates were more likely to meet PrEP eligibility criteria compared to those who did not consider themselves candidates (aRR 1.65; 95% CI 1.34-2.03). Those who had engaged in group or transactional sex were more likely to have behavioral indications for PrEP (aRR 1.27; 95% CI 1.07-1.50, aRR 1.32; 95% CI 1.13-1.56, respectively), whereas HIV test history was not significantly associated with behavioral indications for PrEP. Respondents who had engaged in group sex or transactional sex were more likely to perceive themselves as candidates for PrEP (aRRR 2.24; 95% CI 1.21-4.16, aRRR 2.58; 95% CI 1.09-6.13, respectively), although those never tested for HIV were less likely to perceive themselves as candidates for PrEP (aRRR 0.18; 95% CI 0.03-0.91). The elucidation of candidacy perceptions and risk behaviors is key to furthering the effective implementation of PrEP engagement interventions.