This thesis contributes to the understanding of young people's sexual behaviour, the nonnative context in which it takes place, and how implicit assumptions contained in interventions inadvertently contribute to young people's sexual risk-taking. Fieldwork was carried out in early 2000 in Maputo, Mozambique and consisted of 71 individual and 21 group interviews with men and women aged 16-24 years old. Script theory was used as an organising framework to show how sexual culture and the social construction of risk influence the behaviour change process. Nearly all young people interviewed were sexually active and most reported involvement in a relationship with a long-standing partner. Casual sex was common; however condom use was inconsistent with all partners. The immediate concern for young people is pregnancy, not HIV/AIDS. Two locally defined sexual identities were found. First, the survivor lifestyle demonstrates how gender and power relations define and control sexual identity and determine sexual choices among young people. Second, the saca cena (one-night stand) illustrates how and why some young people come to terms with risk, redefine their sexuality, and adopt innovative sexual behaviour including condom use. Without a clear understanding of how behaviour change happens and how meaning becomes vested in practice, interventions will fail to promote healthy sexual lifestyles and prevent HIV/ AIDS and STls among young people. The gap between knowledge and practice is a case in point. Young people bridge this gap, but their response is mediated by the social and sexual identities they assume, the perceived consequences of mitigating actions, and the resources at their disposal. For the purpose of structuring effective interventions, it is important to assess what these perceived risks are, how they become integrated into sexual identities, and what resources can be deployed to modify them.