Objective: Pulmonary rehabilitation in patients with COPD has been shown to be beneficial but the optimal setting is not known. In the present study, the efficacy of a short-term community-based exercise programme was compared with a standard hospital outpatient programme. Additionally, the usefulness of community or home programmes in maintaining improvements in the longer term was studied.Methodology: Forty-three patients with moderate to severe COPD were randomized to one of the following three groups: a 3-month hospital programme then a 9 month home programme (Hospital/ Home); a 3-month hospital programme then a 9-month community programme (Hospital/ Community); or a 12-month community programme (Community/Community). The initial 3-month programme was analysed by comparing the Hospital group (Hospital/Home plus Hospital/ Community) with the Community group (Community/Community). Six-minute walking distance (6MWD), quality of life (Guyatt chronic respiratory disease questionnaire, CRQ) and lung function were measured at 0, 3, 6 and 12 months and results were analysed using the Wilcoxon rank sum test.Results: At 3 months, there was a significant improvement from baseline in 6MWD in the Hospital group (81.3 +/- 18.3 m, P < 0.05, ANOVA) but not the Community group (14.4 +/- 28.5 m, not significant). The difference between the groups was not significant (P = 0.058). At 3 months, there was a significant improvement in quality of life in the Hospital group (CRQ + 16.3 +/- 3.1, P < 0.01, ANOVA) and in the Community group (CRQ + 10.2 +/- 4.9, P < 0.05, ANOVA) but the difference between the groups was not significant. Following the initial 3-month programme, the dropout rate was high overall (73% by 12 months), and therefore data from the maintenance programme could not be analysed.Conclusions: A 3-month community-based exercise programme for patients with COPD did not improve 6MWD. The long-term retention rates in the programmes were poor.