Objectives To investigate the effect of pharmacy-based medication reconciliation on the frequency of unintentional medication discrepancies in acutely admitted individuals aged 65 and older. Design Multicenter intervention study with pre-post design. Setting Twelve Dutch hospitals. Participants One thousand five hundred forty-three individuals aged 65 and older with an acute hospital admission through the emergency department. Measurements The intervention consisted of the Best Possible Medication History (BPMH), based on combining information from the community pharmacy record, the information provided by a structured interview with participants about their medication use, and medication containers. In nine hospitals, pharmacy technicians obtained the BPMH, and in three hospitals, a mixed model was used (physicians or pharmacy technicians obtained the BPMH). Primary outcome measure was the proportion of participants with one or more unintentional medication discrepancies. The primary outcome measure was stratified according to type of intervention (pharmacy based vs mixed model). Results The proportion of participants with one or more unintentional medication discrepancies was reduced from 62% to 32% [odds ratio (OR) = 0.29, 95% confidence interval (CI) = 0.23-0.37]. These results remained statistically significant after adjustment for type of department and hospital (OR = 0.20, 95% CI = 0.15-0.26), and this effect remained stable for 6 months. Stratified analysis showed that no effect from the intervention was evident in the three hospitals with a mixed-model intervention, in contrast to the hospitals with a pharmacy-based intervention. The medication discrepancy types "omission" and "dosage or strength" occurred most frequently and were the main types that the intervention influenced. Conclusion Pharmacy-based medication reconciliation leads to a substantial reduction in medication discrepancies in acutely admitted elderly adults. © 2013, The American Geriatrics Society.