During the past five years, researchers from the Medical Research Council and Uganda Virus Research Institute (MRC/UVRI) Programme on AIDS have studied sexual behaviour to better understand the risk and the spread of HIV infection in a rural Ugandan community. This paper aims at a reflective critique of the application of various methods of studying sexual behaviour in a series of six studies within the programme. The objectives of these various studies have been different: ranging from the natural history of HIV-infection to marital instability to household coping. This variety of foci has led to multiple research strategies. Three methodological factors influencing the research and the results were identified: the research model; the meanings of research questions; and personal factors affecting the interview relationship. Although the impact of these factors could not be entirely eliminated, precautions could be taken to diminish these biases. Comparing data obtained through different methods proved useful not only as a validity test but also as a mean to more deeply interpret the data according to culture, linguistics and society. Lessons learned during this piece of work include the importance to the quality of data by inviting local communities to participate in the research process; broadening the field of sexuality from a health-oriented model to reach an anthropological perspective; considering the influence of research organization on the context in which sexual behaviour takes place as a part of the study objectives and promoting an inter-disciplinary dialogue overcoming dogma and prejudices.