Thrombolysis is an attractive but potentially dangerous they for cerebral ischemia: it is capable of dissolving an arterial thrombus, but can also transform a pale infarct into a hematoma and/or may cause severe oedema and herniation. The safety and efficacy of the treatment critically depend on the timing of intervention ad on patient selection. In recent studies on ischemic stroke, spontaneous hemorrhagic transformation of an infarct seems to be related to the size of the lesion, and can be reliably predicted as early as five hours from stroke onset by the presence of focal hypodensity in the CT scan. That is why in the European Co-operative Acute Stroke (ECASS), a randomised, double blind trial on intravenous rt-PA in hemispheric stroke, patients showing, on the admission CT scan, extended early hypodensity, involving more than one third of the territory of the middle cerebral artery, were excluded from the day. Other ongoing trials on thrombolytic agents are expected to provide further indications on how to identify those patients most likely to benefit and least likely to experience adverse effects from this treatment.