This study examined whether subacute stereotactic evacuation of basal ganglia haematomas in primarily non-comatose patients is suitable to improve the ultimate outcome of this subgroup of stroke patients. Applying rigorous selection criteria, 56 consecutive non-comatose patients with ganglionic haematomas were treated stereotactically, and 1-year outcomes employing four outcome parameters commonly used to assess outcome were compared with those of 39 similar patients who were treated purely medically. No survival benefit was found in long-term follow-up for either surgical or conservative treatment (total mortality 16.1% vs 28.2%; P=0.121). Among survivors, however, outcome was significantly better in surgical patients. Compared with medical patients, the median Glasgow Outcome Scale score was 1 point higher (P<0.0001) in surgical patients, and the median European Stroke Scale score improvement from baseline to 1-year score was significantly better (P<0.0001). Accordingly, the median Barthel Index score was significantly higher (P=0.002), and the median Modified Rankin Scale score was 1 point lower (P<0.0001). We conclude that primarily non-comatose patients with basal ganglia haematomas can ultimately profit from this form of minimally invasive treatment.