BACKGROUND: There is a high risk of drug-related problems in nursing homes due to polypharmacy, multi-morbidity and age-related changes. We describe the drug use and compare the pharmacotherapy in two nursing homes with different staffing of physicians. MATERIAL AND METHODS: We included 48 long-term patients from two nursing homes in Oslo; i.e. nursing home A (24 patients) and nursing home B (24 patients). A pharmacist recorded information on patients' drug use, identified and classified drug-related problems, and classified the drugs used according to their anticholinergic burden. Two physicians (with experience in geriatrics and nursing home medicine) assessed the clinical importance of the drug-related problems independently from each other. The physicians were blinded with respect to which nursing home the patients came from. RESULTS: Patients in nursing home A used a median (interquartile range [IQR]) of 7.0 (5.3-11.0) drugs, and those in nursing home B used 9.5 (8.0-12.8); the median difference was 2.0, 95% CI 1.0-4.0, p = 0.006). Patients also had lower anticholinergic drug scores in nursing home A (1.0 [0.0-2.0]) than in nursing home B (2.0 [2.0-3.8]); median difference 1.0, 95% CI 0.0-2.0, p = 0.009). Patients in home A also had lower numbers of drug-related problems (3.0 [2.0-4.0]) than those in home B (5.5 [3.3-8.0]); median difference 1.0, 85% CI 0.0-3.0, p = 0.007. No significant differences were found between the nursing homes with regard to patients' age, co-morbidity, kidney function, or dementia state, but nursing home A had a better staffing of physicians. INTERPRETATION: The number of drugs used as well as the quality indicators varied considerably between the nursing homes assessed. Differences in physician staffing might be one reasonable explanation. Our study highlights the importance of systematic multidisciplinary medication reviews for quality improvement in nursing homes.