Summary Background The antiplatelet drug cilostazol is efficacious for prevention of stroke recurrence compared with placebo. We designed the second Cilostazol Stroke Prevention Study (CSPS 2) to establish non-inferiority of cilostazol versus aspirin for prevention of stroke, and to compare the efficacy and safety of cilostazol and aspirin in patients with non-cardioembolic ischaemic stroke. Methods Patients aged 20–79 years who had had a cerebral infarction within the previous 26 weeks were enrolled at 278 sites in Japan and allocated to receive 100 mg cilostazol twice daily or 81 mg aspirin once daily for 1–5 years. Patients were allocated according to a computer-generated randomisation sequence by means of a dynamic balancing method using patient information obtained at registration. All patients, study personnel, investigators, and the sponsor were masked to treatment allocation. The primary endpoint was the first occurrence of stroke (cerebral infarction, cerebral haemorrhage, or subarachnoid haemorrhage). The predefined margin of non-inferiority was an upper 95% CI limit for the hazard ratio of 1·33. Analyses were by full-analysis set. This trial is registered with ClinicalTrials.gov , number NCT00234065. Findings Between December, 2003, and October, 2006, 2757 patients were enrolled and randomly allocated to receive cilostazol (n=1379) or aspirin (n=1378), of whom 1337 on cilostazol and 1335 on aspirin were included in analyses; mean follow-up was 29 months (SD 16). The primary endpoint occurred at yearly rates of 2·76% (n=82) in the cilostazol group and 3·71% (n=119) in the aspirin group (hazard ratio 0·743, 95% CI 0·564–0·981; p=0·0357). Haemorrhagic events (cerebral haemorrhage, subarachnoid haemorrhage, or haemorrhage requiring hospital admission) occurred in fewer patients on cilostazol (0·77%, n=23) than on aspirin (1·78%, n=57; 0·458, 0·296–0·711; p=0·0004), but headache, diarrhoea, palpitation, dizziness, and tachycardia were more frequent in the cilostazol group than in the aspirin group. Interpretation Cilostazol seems to be non-inferior, and might be superior, to aspirin for prevention of stroke after an ischaemic stroke, and was associated with fewer haemorrhagic events. Therefore, cilostazol could be used for prevention of stroke in patients with non-cardioembolic stroke. Funding Otsuka Pharmaceutical.