Objective: Physical activity can reduce the risk of several chronic diseases. Such diseases are most prevalent in economically-disadvantaged groups where physical activity levels are consistently lower. There is a need to engage disadvantaged groups in programmes to increase physical activity. This case study examined programmes on offer in a deprived area and the characteristics of those who participated. Design: Case study. Setting: The setting for this study was Southmead, a highly deprived suburb of Bristol with the city's lowest life expectancy (75.3 years). Method: Desk research and venue visits were used to assess provision of physical activity sessions. A questionnaire covering demographics, attendance patterns and communications channels was completed by 152 adult and adolescent session participants. Results: In the study area, 37 activity sessions were offered. Only 45.4 per cent of attendees were resident in the study area. A mean of 17.5 attended community generated sessions, 22.6 special interest sessions and 6.7 local authority/trust classes. Locals were more likely to attend community generated sessions (49.3 per cent; 87.7 per cent of men) than those provided by the local authority (34.8 per cent; 13.3 per cent of men) (chi[superscript 2] = 27.8, df = 2, P less than 0.001). Word of mouth was the most common mechanism for building session awareness (local 57.8 per cent; non-local 62.9 per cent). Conclusions: Local authority driven physical activity sessions in low-income communities have low levels of participation. Activities with roots in the community are most effective at attracting local participants, particularly men. The most common communication mechanism is word of mouth but whether this is due to lack of investment and expertise in other promotional techniques is not clear. (Contains 2 tables.)