BACKGROUND: A cohort of patients utilizing ?-agonist medications was anonymously identified utilizing BC Pharmacare data. To enable recruitment of a sample of this cohort into a prospective study of The Determinants of Excessive ?-agonist Use , a recruiting strategy was developed to facilitate patient contact/enrollment without violating confidentiality guidelines. METHODS: The study cohort includes patients 19 50 yr. living in the Greater Vancouver Regional District prescribed a short-acting ?-agonist in 1998. Using encrypted personal health numbers (PHNs), each patients primary prescribing physician will be identified, while remaining blind to the patients identity. A random sample of eligible subjects selected for enrollment into the study will then be mixed ( camouflaged ) with encrypted PHNs of patients not receiving a ?-agonist, identified from the Pharmacare database, at a minimum ratio of one patient per physician. PHNs will be then unencrypted, patient addresses obtained, and recruitment letters mailed to patients ( -agonist users and non-users) through their physician, who will be asked to forward the letters to asthmatic patients identified from their practice. RESULTS: 9,420 subjects were identified as eligible for recruitment; 7,056, 1,431, 653, and 280 in the low, intermediate, high and excessive use groups, respectively received prescriptions from 3,039 physicians. 500 patients will berandomly selected by Pharmacare from the low, intermediate, and high use groups; all patients in the excessive use group will be recruited. A primary prescribing physician will be identified for each patient, and at least one non- -agonist user will be selected for each physician. Recruitment letters will be sent to 1,780 patients, through their physicians, in an effort to enroll 520 asthmatics. CONCLUSIONS: This recruiting methodology has been approved by two independent ethics committees. Utilizing anonymous patient data and camouflaged sampling will ensure the protection of patient privacy while facilitating the recruitment of a representative, stratified sample of the study population.