OBJECTIVE: To report young people's variations in sexual health knowledge, attitudes and behaviours by religious affiliation. DESIGN: A cross-sectional, questionnaire-based survey administered in 16 Secondary/High schools in London, UK. The sample consisted of 3007 students in school Years 11-13 (aged 15-18), present in school on the day of questionnaire administration. Excluding those who described themselves as 'Other religious', 15.9% (n=418) described themselves as having no religious affiliation, 36.3% (n=957) were Christian, 25.8% (n=679) were Muslim and 22.1% were Hindu (n=582). The religious affiliation varied within and across different ethnic groups. METHOD: Self-administered questionnaire completed under 'exam' conditions, either in tutor groups or a school hall. The 30-minute questionnaires were distributed and collected by a team of ethnically and religiously diverse fieldworkers. RESULTS: Religious students, as opposed to those reporting no religious affiliation, generally reported poorer sexual health knowledge, and were more conservative in their attitudes to sex. Among males and females, those with no religious affiliation and Christian students reported the highest prevalence of sexual intercourse by some margin (around 20 percentage points) over the Hindus and Muslims. Christian males most frequently reported sexual intercourse at 49.7%, and Muslim females the least at 9.0%. Among those reporting sexual intercourse, risk behaviours among all religious and non-religious students were evident. Over one-third of Muslim females who had sexual intercourse did not use contraception on their first occasion compared to 10% of those with no religious affiliation, 12% of Christians and 20% of Hindus. Christian and Muslim females reported the highest prevalence of ever not using contraception at 55%, and non-use of contraception with two or more sexual intercourse partners at 14%. CONCLUSION: The findings demonstrate diverse sexual health knowledge, sexual attitudes and sexual behaviours among young people with different religious affiliations. These variations demonstrate the importance of tailoring health education and promotion interventions to meet the specific needs of young people from a variety of different religions. The challenge ahead is to find ways to work with these young people to broach such sensitive issues.