BackgroundSmoking rates in south Asian communities in the UK, particularly men of Pakistani and Bangladeshi origin, are higher than in the general population. As a result, second-hand smoking is widespread in households, which poses a serious health hazard to non-smokers, especially children. Reductions in smoking within the household can help to lower exposure to second-hand smoke. We adapted the smoke-free homes (SFH) educational programme to use Muslim faith leaders as educators and advocates in an attempt to encourage Pakistani and Bangladeshi communities to negotiate and implement smoking restrictions within their homes. We have developed a structured educational guide for Muslim faith leaders. Our primary aim was to assess whether mosques could be acceptable and feasible settings for quality health-promotion programmes. Additionally, we aimed to develop preliminary programme theory for the SFH intervention that could be adapted to other settings and ethnic groups. MethodsThe intervention package was developed from exploratory qualitative research and participatory workshops done with potential users. We tested the programme in five mosques in Leeds, UK. This test was followed by qualitative assessment, with interviews and focus group discussions (FGDs). We did ten semistructured interviews with Muslim religious teachers and chairmen of mosque committees. Four focus groups, each with between five and eight participants, were undertaken with adults attending the mosques or associated women's circles in which the intervention was implemented. We recorded interviews and FGDs and transcribed them verbatim. We analysed data thematically. Three transcripts were translated from Urdu to English and coded by three researchers (two who speak Urdu). A coding frame was then agreed and the remaining transcripts were coded by the two Urdu-speaking researchers. Codes were both a priori (based on themes we had emphasised in the question guides) and emergent. We identified participants in FGDs with the help of religious leaders. Therefore, there could be a correlation between agreeing to deliver the intervention (religious leader), participation in the FGDs (participants), and positive feedback. We attempted to address this limitation by selecting groups with participants who had attended SFH sessions only once with regular attendees to better understand both facilitators and barriers to the SFH intervention. These participants were also shown religious leaders' SFH guide and asked to discuss their experience and share their views of the material. FindingsThe findings indicate that religion and religious leaders—such as Imams, Qur'an teachers, and men and women's circle leaders—are both relevant and important drivers of health promotion measures. There are various barriers to and motivators of to health promotion in mosques, including both surface and deep-seated factors such as existing beliefs, assumptions, experiences, associations, and unique cultural and religious constraints that are essential for understanding and development of a programme theory of SFH. For example, Islamic discourse was believed to have priority over deviant behaviours and resulted in acceptance of smoking restrictions in the home. Family members can act as both allies or barriers to the change, which was sometimes overcome through other support networks (ie, mosque or religious leaders). InterpretationThe mosque is an acceptable and feasible setting for high-quality health promotion interventions. What is needed is to further investigate similar community-based approaches and develop innovative solutions to use existing beliefs and community linkages to encourage positive change. Additionally, the capacity of the communities to promote a smoke-free environment within homes should be developed. Investigation of processes involved is fundamental for development of theory-based, real-programmatic models for intervention and policies. FundingUniversity of Leeds and NHS Leeds.