The Lung Health Study (LHS), a multicenter randomized clinical trial of treatment to prevent chronic obstructive pulmonary disease (COPD), recruited, over a 28-month period, 5887 smokers aged 35-60 who had evidence of mild airflow obstruction on pulmonary function testing. The LHS participants were identified from a pool of over 73,000 age-eligible smokers who underwent the initial pulmonary function screening tests. Methods of recruitment in the 10 centers are here classified into five general strategies: worksites, public sites, mail/phone, media, and other. This paper deals with the results of each of these general methods and their effectiveness in producing participants eligible for the study. The most effective strategies proved to be mail/phone and media. Ongoing monitoring and a flexible approach to recruitment characterized the most successful clinics. Over the recruitment period, LHS clinics shifted their emphasis away from worksite and public site strategies and tended to focus on variations of the mail/phone strategy. Female screenees who passed first-screen eligibility criteria were more likely than males to refuse further participation.