Introduction: CME program planners are being asked to move beyond assessments of knowledge to assessing the impact of CME on practice and patient outcomes. Methods: We conducted a pre-post analysis of administrative data from 107 physicians, nurse practitioners (NPs), or physician's assistants (PAs) who attended one or two continuing medical education (CME) programs (an in-person, mainly didactic session on breast complaints in women, or an individual mentorship with general surgeons) between August 2002 and March 2003. We examined associations between the number of trainings and attempted breast mass aspirations or general surgery referrals for breast masses; individual training and breast mass aspiration attempts or general surgery referrals; and provider type and attempted breast mass aspirations. Generalized linear mixed models were used to model dichotomous outcomes. Results: Clinicians who participated in individual trainings performed more breast mass aspirations after training (odds ratio (OR) 3.07, [95% confidence interval 1.10 - 8.54]). Participants who completed two trainings performed more breast mass aspirations after training (OR 2.33, [1.19 - 4.57]), while those who completed just one did not (OR 1.34, [0.39, 4.58]) but the effect started after the first training and did not strengthen after the second training. NPs and PAs attempted more aspirations after training (OR 6.1, [1.54, 24.1]), whereas physicians did not (OR 0.89 [0.36, 2.22]). Training was not associated with a change in referrals to general surgery. Referral appropriateness, pre-training readiness to change, and previous training in breast mass aspiration were not assessed. Discussion: Attempts to aspirate breast masses may increase after CME training. Individual training may be more effective than group training in increasing the likelihood of attempted aspirations.