Background: Recruitment to clinical trials is often slow and difficult, with a growing body of research examining this issue. However there is very little work related to stroke. Aims: The aim of this study was to examine the success and efficiency of recruitment of community-dwelling stroke survivors over the first two years of a clinical trial aiming to improve community ambulation. Method: Recruitment strategies fell into 2 broad categories: (i) advertisement (such as newspaper advertising and media releases), and (ii) referral (via hospital and community physiotherapists, a stroke liaison officer and other researchers). Records were kept of the number of people who were screened, were eligible and were recruited for each strategy. Results: The recruitment target of 60 in the first two years was not met. 111 stroke survivors were screened and 57 were recruited (i.e., a recruitment rate of 51%). The most successful strategy was referral via hospital-based physiotherapists (47% of recruited participants) and the least successful were media release and local newspaper advertising. The referral strategies were all more efficient than any of the advertisement strategies. Conclusions: In general, recruitment was inefficient and costly in terms of human resources. Given that stroke research is underfunded, it is important to find efficient ways of recruiting stroke survivors to clinical trials. An Australian national database similar to other disease-specific data bases (such as the National Cancer Database) is under development. In the interim, recruiting for several clinical trials at once may increase efficiency.