Intracranial hemorrhage is a critical factor when considering efficacy and safety of thrombolytic intervention after thromboembolic stroke. This study tested whether magnetic resonance imaging could identify tissue for hemorrhagic transformation after the onset of embolic stroke. Rats subjected to embolic stroke with and without recombinant tissue-type plasmogen activator (rt-PA) treatment were followed-up with magnetic resonance imaging using the inverse of the apparent forward transfer rate for magnetization transfer (k(inv)), gadolinium-chelate contrast-enhanced magnetic resonance imaging, and diffusion-, perfusion-, and T2-weighted imaging. Rats with embolic stroke either were treated with rt-PA 1 (n = 16) or 4 hours (n = 13) after stroke onset or were not treated (n = 15). From these groups, at total of 17 rats had intracerebral hemorrhage. Tissue progressing to hemorrhage and adjacent to the site of hemorrhage was analyzed to identify magnetic resonance imaging markers that characterize hemorrhagic transformation. The parameter maps of k(inv) and contrast-enhanced magnetic resonance imaging showed greater sensitivity in the detection of tissue destined for hemorrhagic transformation compared with the apparent diffusion coefficient of water (ADCw) and CBF. In tissue not destined to undergo hemorrhagic transformation, k(inv) maps and contrast-enhanced magnetic resonance imaging exhibited small increases in k(inv) and contrast-enhanced magnetic resonance imaging signal intensity in the area encompassing the territory supplied by the middle cerebral artery. In contrast, large increases in k(inv) and in signal intensity in the contrast-enhanced magnetic resonance images were detected in the region where gross hemorrhage was confirmed histologically. The values of k(inv), T2, and signal intensity in the contrast-enhanced magnetic resonance images were significantly higher in the region destined for hemorrhagic transformation (k(inv), P < or = 0.033 3-24 hours after embolization; T2, P < or = 0.037 24-48 hours; contrast-enhanced magnetic resonance imaging, P < 0.05 4-7 hours) compared with the nonhemorrhagic transformation ischemic region or in the contralateral homologous regions after onset of ischemia. Of these methods, k(inv) shows the most sensitivity in the detection of hemorrhagic transformation soon after embolization. The authors' data suggest that k(inv) and contrast-enhanced magnetic resonance imaging are potentially important methodologies for detecting tissue destined for hemorrhagic transformation.