Introduction: new infections of Human Immunodeficiency Virus (HIV) remain a big problem in many countries. Different interventions have been implemented to improve the general knowledge of HIV, with the hypothesis that increasing HIV knowledge will reduce risky sexual behavior (RSB). However, HIV knowledge may not necessarily reduce RSB. This study explores HIV knowledge and its effect on RSB.

Methods: the study used data from the 2005 and 2010 Rwanda Demographic and Health Surveys to analyze the association between HIV risk factors and two types of RSB (having two or more partners in the past 12 months; and among those with two or more partners, not using a condom at last sex) and the association between HIV knowledge and those RSB. Multivariate logistic regression was used to determine predictors of RSB.

Results: among 2,773 men in 2005 and 3,772 men in 2010, 5% and 7% respectively reported having two or more sexual partners. Among them, 93% in 2005 and 74% in 2010 did not use a condom at the last sex. Between 2005 and 2010, knowledge of the protective effect of having just one uninfected faithful partner, and basic knowledge of HIV decreased. Knowledge of the protective effect of using condoms increased from 90% to 94%. However, HIV knowledge was not associated with either type of RSB.

Conclusion: in setting up policies and strategies related to HIV prevention, policymakers should consider that focusing on HIV knowledge is not sufficient in itself.