Introduction: Inappropriate prescribing is a prevalent problem in nursing home residents that is associated with cognitive and physical impairment. Few interventions have been shown to reduce inappropriate prescribing. The aim was therefore to examine successful discontinuation of inappropriate medication. Methods: A cluster randomized controlled trial was conducted. Fifty-nine wards were randomly assigned to the intervention or to "care as usual". The intervention was a Multidisciplinary Multi-step Medication Review (3MR), consisting of an assessment of the patient perspective, a medical history, a critical appraisal of medication, a meeting between the elderly care physician and a pharmacist, and the execution of medication changes. The primary outcome was successful discontinuation of ≥1 inappropriate drug(s), without relapse or severe withdrawal symptoms. Secondary outcomes included neuropsychiatric symptoms, cognitive function and quality of life. Nursing home residents with a life expectancy of >4 weeks who did not refuse treatment with medication were included. Data were collected at baseline and at an average follow-up of 144 days. Results: A total of 426 nursing home residents participated (intervention group: N=233 and control group: N=193). Generalized linear mixed models (logit link function) showed that for 91 (39.1%) of the residents in the intervention group ≥1 inappropriate drugs could be successfully discontinued vs. 57 (29.5%) of residents in the control group (adjusted odds-ratio: 1.57, 95% CI: 1.03 to 2.39). There was no deterioration on secondary outcomes. Conclusions: The 3MR is effective in discontinuing inappropriate medication in nursing home residents whilst probably not compromising their wellbeing.