In a retrospective follow-up study 38 patients with aneurysmal subarachnoid haemorrhage (SAH) and 20 patients without an angiographically proven source of SAH were tested neuropsychologically one to five years after the acute event. All patients were operated on early within 72 hours if an aneurysm was proven angiographically and all were treated with nimodipine. Both patient groups had comparable cognitive deficits in spite of the less severe SAH of non-aneurysmal origin. Only in two cognitive functions the groups differed significantly. The patients after non-aneurysmal SAH had a significantly lower mean in the IQ subtest similarities finding (p < 0.05), while the patients after aneurysmal SAH had a significantly lower mean in a test of visual cognition (p < 0.05). A more detailed analysis with clinically homogeneous subgroups was additionally performed. The results showed in the subgroup with poor clinical grades that patients with aneurysmal SAH were significantly more disturbed in focal cognitive functions like short- and long-term memory and word-finding capacity, while patients with SAH of unknown origin scored significantly worse in a neuropsychological test related to attention, which can be regarded as a more diffuse cognitive function.