The purpose of this study was to develop a method that eliminates the influence of the T1 relaxation time upon the signal-time course in perfusion-weighted imaging of cerebral lesions with blood-brain-barrier (BBB) disruption. On a 1.5 T whole body clinical magnetic resonance (MR) imager, we implemented a dual-echo RF-spoiled FLASH sequence (TE=6/23.6 ms). We developed a postprocessing routine that allowed to calculate a signal-time course representing only the change in T2* and another one representing only the change in T1. Using this method, we examined 7 patients with various brain lesions showing evidence of BBB disruption. In the signal-time-curves obtained from the early echo we found a distinct signal drop due to the T2* effect. These effects could be eliminated by the correction algorithm yielding a 67% higher signal increase. Correction of the signal-time curve of the late echo yielded a more pronounced maximum signal drop and a decrease in postcontrast signal intensity. We found that without this correction the relative regional cerebral blood volume and the first moment of the concentration-time curve were underestimated by 72% and 22%, respectively. The dual echo-sequence combined with the postprocessing algorithm separates T1 and T2* effects and thus allows to assess cerebral hemodynamics and contrast agent kinetics simultaneously. This method may be a useful tool for characterizing, staging, and therapy monitoring of brain tumors.