Participant recruitment is often the rate-limiting step in clinical trials, but the methods used and the processes involved in recruitment are rarely reported. Although various methods will often be adopted, there is a dearth of information about the factors which facilitate or impede recruitment to trials. Such information is potentially very valuable to researchers. The present pilot study aimed to recruit a minimum of 30 participants through primary care to a randomized controlled trial (RCT) of three interventions for older adults with depression. The aim was to determine how recruitment to RCTs in primary care may be optimized. A wide range of recruitment methods was used, including both direct general practitioner (GP) referrals and self-referrals. Potential participants were informed about the study in a variety of ways, including GP recruitment when attending surgeries, contacting suitable patients identified from the GP database, posters and leaflets in the GP surgery, and the specific targeting of older people by the distribution of leaflets at 'flu clinics. Both GP referrals and direct referrals were successful. Interestingly and rather unexpectedly, the greatest numbers of individuals were recruited by self-referral. No differences were found between these two groups as to the severity of their depression. Recruiting from primary care can be successful, if primary health care teams are willing to work in partnership with researchers, and if researchers are willing to be flexible and try many approaches. In particular, the study found that cultivating self-referrals was a profitable strategy, which was not what had been initially expected with this particular target group of older people with depression. These self-referrals were a cost-effective method of recruitment, each costing around one-third less than those recruited by other methods, and without them the recruitment targets would not have been met.