OBJECTIVE: This study aimed to evaluate factors associated with inconsistent condom use with casual partners in a population of men who have sex with men (MSM) in Brazil. METHODS: In 2016, 4,176 MSM >18 years were enrolled in 12 capitals of Brazil using a Respondent Driven Sampling (RDS) method. For the construction of the outcome, we evaluated questions about condom use in all anal intercourse (receptive and insertive) in the previous six months and the last sexual intercourse. Estimates were calculated using a weighted complex sample design. We performed a logistic regression analysis to determine the associations between sociodemographic and behavioral factors and inconsistent condom use in sexual relationships with casual male partners. RESULTS: More than half of our sample (50.8%) had not used condoms consistently with casual partners in the previous six months. Inconsistent condom use was significantly associated with: low education (weighted odds ratio - wOR: 1.55; 95% confidence interval - CI 0.99-2.40), lack of counseling on sexually transmitted infections STI (wOR: 1.51; 95%CI 1.05-2.17), non-use of condoms at sexual debut (wOR: 3.05; 95%CI 2.12-4.40) and moderate and high perceived risk for HIV (wOR: 1.51; 95%CI 1.07-2.14). Higher age was negatively associated with inconsistent condom use (wOR=0.97, 95%CI 0.89-0.99). CONCLUSION: Despite being an individual behavior, condom use is related to factors beyond the individual scope. HIV/Aids prevention policies should focus on younger MSM, providing qualified information about condom use, preferably before the beginning of their sexual life.