The effectiveness of a pharmacy-obtained medication history on the medication reconciliation process in the behavioral health unit (BHU) of a community hospital was studied. Patients admitted to the BHU of a 411-bed, not-for-profit hospital from 6 a.m. on Monday through 12 p.m. on Friday from September 1, 2005, through October 6, 2005, were candidates for the study. Within 18 hours of admission to the BHU and after the medication history had been obtained by a nurse, a pharmacy technician gathered patient demographic and medication information from the chart and the patient's medication bottles. Once the technician completed the documentation, the pharmacist was notified of a new admission. The pharmacist reviewed the collected documentation and patient chart before interviewing the patient. Of the 54 patients who met the study's inclusion criteria, 91% were seen by a pharmacist within 18 hours of admission. The mean +/- S.D. time delay to interview the patient was 11.6 +/- 5.1 hours. Pharmacists spent a mean of 13.9 minutes completing patients' medication histories. The mean +/- S.D. number of medications identified by nursing on admission was 4.0 +/- 3.2, compared with 5.3 +/- 3.7 identified by pharmacists (p < 0.05). The mean number of medication discrepancies identified per patient was 2.9. Of the discrepancies, 48% were related to an omitted or incorrect medication, 31% to an omitted or incorrect dose, and 13% to an omitted or incorrect frequency; 8% were categorized as miscellaneous. Pharmacists' participation in obtaining patients' medication histories through chart review and patient interview increased the effectiveness of the medication reconciliation process in an inpatient BHU.