Stroke due to basilar artery occlusion has a high mortality and morbidity. Intra-arterial thrombolysis has been reported to improve survival and outcome status. Our aim was to assess the safety and efficacy of intra-arterial urokinase in a consecutive series of patients with clinically severe brainstem ischaemic stroke and major vertebrobasilar vessel occlusion. Incremental doses of urokinase were administered until clot lysis was achieved, or until a limit of 1,000,000 U. Patients were then anticoagulated with heparin and warfarin, and 6-month functional status was measured by the Barthel index. Sixteen patients, aged 22–73 (median 60), were treated 5–31 (median 15) h following symptom onset. Thirteen of the 16 patients (82%) had initial complete or partial recanalisation. Complete occlusion of the basilar artery was present in 13, and recanalisation was achieved in 10 of these (77%), although 2 re-occluded. Four of 5 patients with persistent occlusion died, compared with only 1 death in 8 patients with sustained recanalisation (p = 0.02, Fisher''s exact test, one-tailed). Intra-arterial urokinase can recanalise basilar artery occlusion, with significant reduction in mortality at 6 months. A prospective randomised, controlled trial is necessary to confirm the benefit of this therapy.