Recent studies have raised concerns about the safety of low frequency ultrasound in transcranial therapeutic application in cerebral ischemia. This study was designed to evaluate safety aspects and potential deleterious effects of low frequency, 60 kHz ultrasound in treatment of experimental middle cerebral artery occlusion (MCAO) in rats. Forty-five male Wistar rats were submitted to either temporary (90 min; groups I and II) or permanent MCAO (groups III and IV) using the suture technique. All animals received recombinant tissue plasminogen activator (rt-PA) starting 90 min after the beginning of occlusion. Groups I and III were additionally treated with 60 kHz ultrasound (time average acoustic intensity 0.14 W/cm(2), duty cycle 50%). Outcome assessment consisted of magnetic resonance imaging (MRI) and clinical evaluation after 5 and 24 h, and histology (perfusion fixation after 24 h). Overall mortality was higher in animals treated with ultrasound (43% versus 29% in controls). Most animals died during the insonation period (25% in group I, 36% in group III, no animals in the corresponding control groups; p < 0.05). Histology revealed disseminated microscopic intracerebral bleeding and subarachnoid hemorrhage as one possible cause of death. After temporary occlusion, the hemispheric ischemic lesion volume was more than doubled in animals treated with ultrasound (20.3% +/- 14.1% versus 8.6% +/- 5.1% in controls; p < 0.05). No difference in lesion volume was seen after permanent MCAO. Neurological assessment showed impairment of hearing as an additional specific side effect in ultrasound treated animals (65%, no impairment in controls). Although the results are not directly transferable to the human setting, this study clearly demonstrates the potential limitations of low frequency therapeutic ultrasound and the importance of pre-clinical safety assessment.