Female genital cutting (FGC) is a widespread cultural practice in Africa and the Middle East, with a number of potential adverse health consequences for women. It was hypothesized by Kun (1997) that FGC increases the risk of HIV transmission through a number of different mechanisms. Using the 2003 data from the Kenyan Demographic and Health Survey (KDHS), this study investigates the potential association between FGC and HIV. The 2003 KDHS provides a unique opportunity to link the HIV test results with a large number of demographic, social, economic and behavioural characteristics of women, including women's FGC status. It is hypothesized that FGC increases the risk of HIV infection if HIV/AIDS is present in the community. A multilevel binary logistic regression technique is used to model the HIV status of women, controlling for selected individual characteristics of women and interaction effects. The results demonstrate evidence of a statistically significant association between FGC and HIV, after controlling for the hierarchical structure of the data, potential confounding factors and interaction effects. The results show that women who had had FGC and a younger or the same-age first-union partner have higher odds of being HIV positive than women with a younger or same-age first-union partner but without FGC; whereas women who had had FGC and an older first-union partner have lower odds of being HIV positive than women with an older first-union partner but without FGC. The findings suggest the behavioural pathway of association between FGC and HIV as well as an underlying complex interplay of bio-behavioural and social variables being important in disentangling the association between FGC and HIV.