Religious affi liation has been associated with the prevalence of the human immune-defi ciency virus (HIV) in India and South Africa. Limited research has explored whether any such association exists in Malawi. The current study, using data from the Malawi Demographic and Health Survey (MDHS) 2004 -2005, was conducted to examine the association among religion, HIV infection, and sexual behaviors among 15-49-year-old women. Logistic regression analysis was used to assess the association. A total of 2609 women were included, of whom 2181 (83.5 % ) were Christians and 407 (15.6 % ) were Muslims, reported ever having sex and had HIV test results within the survey. In multi-variate analysis assessing HIV infection, no differences were observed among Protestants and Catholics, Muslims, and all Christians combined or between Catholics compared with individual religious denominations. The lack of disparity in HIV infection across religious groups was observed despite the fact that there were some differences in sexual behaviors. In the predominantly Christian society of Malawi, where almost all people are affi liated with one religion or the other and Islam is the second largest religion, differences in HIV prevalence by religion was not observed as has been reported in other settings. This lack of disparity was the case although there were some differences in sexual behaviors, such as in premarital sex and condom use at sexual debut, which have potential to infl uence risk of HIV infection. We suggest that the lack of sexual networks ' boundaries defi ned by religion may explain this fi nding.