Recruitment effort and costs in clinical trials are very often underestimated. As funding constraints increase, more precise estimates of costs as well as methods to monitor effectiveness are needed. However, few studies exist that report recruitment cost effectiveness and costs. The study reported here was developed in Portland, Oregon as an ancillary study to the Lung Health Study. The authors developed a monitoring and evaluation system to track response rates and costs associated with each of five recruitment methods. These methods include (1) media (TV, radio, newspapers), (2) neighborhood promotion, (3) direct mail, (4) worksite promotion, and (5) referral by other study participants. The analysis is limited to a 6-month period (April 1-September 30, 1987) or the middle phase of recruitment. During the study period, 46% were recruited from media, 30% from neighborhood promotion, 1.5% from direct mail, 11% from worksites, and 12% from referrals. Neighborhood promotion and direct mail were least cost-effective, media most cost-effective, with worksite and referral moderately cost-effective. The cost-effectiveness of media promotion is explained in part by the limited number of media sources in Portland, making it possible to reach a large audience with less effort, as well as the ability to provide rapid feedback to media sources. We conclude that the effectiveness of recruitment in a clinical trial is maximized by using multiple overlapping recruitment strategies coupled with a monitoring system that can provide rapid feedback regarding the effectiveness and costs of each strategy.