to measure the impact of pharmacist-conducted clinical medication review with elderly care home residents. randomised controlled trial of clinical medication review by a pharmacist against usual care. sixty-five care homes for the elderly in Leeds, UK. a total of 661 residents aged 65+ years on one or more medicines. clinical medication review by a pharmacist with patient and clinical records. Recommendations to general practitioner for approval and implementation. Control patients received usual general practitioner care. primary: number of changes in medication per participant. Secondary: number and cost of repeat medicines per participant; medication review rate; mortality, falls, hospital admissions, general practitioner consultations, Barthel index, Standardised Mini-Mental State Examination (SMMSE). the pharmacist reviewed 315/331 (95.2%) patients in 6 months. A total of 62/330 (18.8%) control patients were reviewed by their general practitioner. The mean number of drug changes per patient were 3.1 for intervention and 2.4 for control group (P < 0.0001). There were respectively 0.8 and 1.3 falls per patient (P < 0.0001). There was no significant difference for GP consultations per patient (means 2.9 and 2.8 in 6 months, P = 0.5), hospitalisations (means 0.2 and 0.3, P = 0.11), deaths (51/331 and 48/330, P = 0.81), Barthel score (9.8 and 9.3, P = 0.06), SMMSE score (13.9 and 13.8, P = 0.62), number and cost of drugs per patient (6.7 and 6.9, P = 0.5) (pounds sterling 42.24 and pounds sterling 42.94 per 28 days). A total of 75.6% (565/747) of pharmacist recommendations were accepted by the general practitioner; and 76.6% (433/565) of accepted recommendations were implemented. general practitioners do not review most care home patients' medication. A clinical pharmacist can review them and make recommendations that are usually accepted. This leads to substantial change in patients' medication regimens without change in drug costs. There is a reduction in the number of falls. There is no significant change in consultations, hospitalisation, mortality, SMMSE or Barthel scores.