OBJECTIVES: To study the effects of two sequential changes in drug cost-sharing policies on adherence to statins and beta-blockers by seniors in British Columbia. METHODS: For each drug class, we identified a baseline cohort of subjects initiating therapy in the 6 months prior to January 1, 2001, a co-payment cohort initiating therapy in the 6 months prior to the co-payment policy introduced January 1, 2002, and a co-insurance cohort initiating therapy in the 6 months prior the co-insurance policy introduced May 1, 2003. We calculated the proportion of patients adherent in each cohort each month, with follow-up for each cohort beginning at the start of that cohort's recruitment period and ending 15 months later. Patients were defined as adherent during a month if they had a proportion of days covered (PDC) of 80% or greater, calculated by dividing the number of days the patient had drug supply available by the number of cohort membership days the patient contributed in that calendar month. RESULTS: In the baseline cohort, which did not experience cost-sharing, 55.8% of statin initiators were adherent to their statins at month 15. The adherence level in the co-payment cohort at this time, 9 months after the introduction of the co-payment policy, was 50.5%. 50.8% of co-insurance cohort member were adherent. Adherence to beta-blockers was lower, with 48% of the baseline cohort initiators adherent at month 15. However, the introduction of the co-payment and co-insurance policies reduced this adherence level by only 1 percentage point. CONCLUSION: The introduction of the co-payment and co-insurance policies reduced adherence to statins by 5 percentage points relative to baseline levels, but had a much smaller effect on beta blocker adherence levels. Policy-makers should consider drug cost as a potential modifier of policy effects.