PURPOSE: To determine whether practicing physicians receiving only clinical information at a traditional continuing medical education (CME) lecture (control group) and physicians receiving clinical information plus information about barriers to behavioral change (study group) would alter their clinical behaviors at the same rate. METHOD: In a randomized controlled trial, the investigators matched 13 pairs of U.S. and Canadian medical schools, assigning one school from each pair to study or control conditions. Following the commitment-to-change model, the investigators asked the primary care physicians attending control or study lectures on the management of cardiovascular risks whether they intended to make behavioral changes as a result of participating in the lectures and, if so, to indicate the specific changes. Thirty to 45 days later, the investigators surveyed the responding physicians to learn whether they had implemented those changes. RESULTS: Information about barriers to change did not increase the likelihood that physicians in the study group would report successful changes; they were no more likely to change than those in the control group. However, the physicians in both study and control groups were significantly more likely to change (47% vs 7%, p < .001) if they indicated an intent to change immediately following the lecture. CONCLUSIONS: Successful change in practice may depend less on clinical and barriers information than on other factors that influence physicians' performances. To further develop the commitment-to-change strategy in measuring the effects of planned change, it is important to isolate and learn the powers of individual components of the strategy as well as their collective influence on physicians' clinical behaviors.