Background: Smoking prevalence is high among Pakistani and Bangladeshi men in the UK, but there are few tailored smoking cessation programmes for Pakistani and Bangladeshi communities. The aim of this study was to pilot a cluster randomised controlled trial comparing the effectiveness of Pakistani and Bangladeshi smoking cessation outreach workers with standard care to improve access to and the success of English smoking cessation services. Methods: A pilot cluster randomised controlled trial was conducted in Birmingham, UK. Geographical lower layer super output areas were used to identify natural communities where more than 10% of the population were of Pakistani and Bangladeshi origin. 16 agglomerations of super output areas were randomised to normal care controls vs. outreach intervention. The number of people setting quit dates using NHS services, validated abstinence from smoking at four weeks, and stated abstinence at three and six months were assessed. The impact of the intervention on choice and adherence to treatments, attendance at clinic appointments and patient satisfaction were also assessed. Results: We were able to randomise geographical areas and deliver the outreach worker-based services. More Pakistani and Bangladeshi men made quit attempts with NHS services in intervention areas compared with control areas, rate ratio (RR) 1.32 (95%CI: 1.03-1.69). There was a small increase in the number of 4-week abstinent smokers in intervention areas (RR 1.30, 95%CI: 0.82-2.06). The proportion of service users attending weekly appointments was lower in intervention areas than control areas. No difference was found between intervention and control areas in choice and adherence to treatments or patient satisfaction with the service. The total cost of the intervention was 124,000; pound an estimated cost per quality-adjusted life year (QALY) gained of 8,500 pound. Conclusions: The intervention proved feasible and acceptable. Outreach workers expanded reach of smoking cessation services in diverse locations of relevance to Pakistani and Bangladeshi communities. The outreach worker model has the potential to increase community cessation rates and could prove cost-effective, but needs evaluating definitively in a larger, appropriately powered, randomised controlled trial. These future trials of outreach interventions need to be of sufficient duration to allow embedding of new models of service delivery.