Thrombolytic agents are aimed at restoring arterial patency thus reducing the risk of cerebral infarction in case of arterial occlusion by blood thrombus or embolus. A review of data from the literature is presented, and the authors have tried to answer the following questions: 1) What are the probabilities of early mortality (first 30 days) and of morbidity due to untreated or medically (thrombolysis excluded) treated cerebral infarction? 2) What are the morbidity and mortality rates after systemic thrombolysis or selective intraarterial thrombolysis? 3) Are the results of thrombolysis affected by the site of arterial occlusion (carotid or vertebrobasilar system)? 4) Does thrombolysis increase the risk of cerebral haemorrhage? 5) Can thrombolysis be regarded as an effective treatment and, if so, in which cases? This study of systemic thrombolysis yielded a mean mortality rate of 22.2%, with satisfactory functional outcome in 58.2% of the survivors. The corresponding figures for selective intraarterial thrombolysis were 29.7% and 84.5% respectively. However when the arterial territories occluded were taken into account, they became 10.4% and 84.4% respectively in the carotid system and 61.5% and 85.00% in the vertebrobasilar system. Comparisons between the natural history of cerebral infarctions, as far as it is known, and the results of thrombolysis suggest that thrombolytic agents are: 1) probably justified by the selective intraarterial route when the occlusion lies in the carotid system; 2) not indicated by the systemic route with urokinase or streptokinase but trials with tPA are being performed. These conclusions, however, must be interpreted with caution as they rest on general data leaving out a number of factors that are often neglected in reports particularly the exact site of arterial occlusion, chiefly in the vertebrobasilar system. From 3 limited cohorts it was possible to evaluate at about 14% the mortality rate in thrombolysis for basilar artery occlusion. This figure, compared with the 70-80% estimate found in the literature for untreated occlusions, suggests that the selective intraarterial thrombolysis may perhaps be indicated also in basilar artery occlusions. It must be noted that selective intraarterial thrombolysis is still an exceptional treatment which requires sophisticated techniques, highly qualified neuroradiologists and prompt application i.e. emergency stroke units.