Fluid-attenuated inversion recovery (FLAIR) sequence is currently used in clinical practice. Some reports emphasize the possibility that, in pathologic conditions, intravenous injection of gadolinium chelates may lead to an increased signal inside the cerebrospinal fluid (CSF). The aim of this study was to evaluate the presence of CSF signal changes in pathologic conditions causing blood-brain barrier disruption or neovascularization when imaging is performed after intravenous injection of gadolinium. We obtained FLAIR sequences after gadolinium injection from 33 patients affected by different intracranial pathologies and 10 control subjects. Patients were affected by ischemic stroke in the subacute phase, from 2 to 7 days from onset of symptoms (12 patients), meningiomas (8 patients), high-grade gliomas (5 patients), previous surgical procedures for intra-axial neoplasms (5 patients), and multiple sclerosis with active plaques (3 patients). Magnetic resonance imaging was performed in patients and controls using a 1.5-T magnet, using T2- and T1-weighted FLAIR sequences. The FLAIR sequence was acquired before and 1-3 h after injection of a standard dose of gadolinium. In those patients affected by ischemic lesions, FLAIR sequences were repeated the next days and 3-4 days later. The CSF signal was visually evaluated by two readers and scored from 0 to 3 depending by the degree of enhancement. The location of CSF signal changes (close to the lesion, hemispheric, or diffuse) was also considered. The CSF signal was markedly increased after 3 h from intravenous injection of gadolinium in all the patients with stroke, in those with previous surgery, and in those with high-grade gliomas whose neoplasm's surface was in contact with the subarachnoid spaces (SAS) or ventricles; a strong enhancement was also evident inside the necrotic component of the tumor. The CSF changes were more evident close to the pathology and/or in the hemisphere involved by the pathology. Moderate CSF enhancement was observed in the SAS close to meningiomas. No signal changes were evident in all the others. In those patients with stroke imaged in the following days, CSF signal showed to be diffuse to both hemispheres the next day and returned to normal values within 2 days. In patients affected by pathologies with blood-brain barrier breakdown or neovascularization close the SAS or the ventricles, CSF changes, related to gadolinium leakage, are likely when FLAIR sequences are acquired 2-24 h after i.v. injection of the contrast. This pattern should be known in order to differentiate it from that of subarachnoid hemorrhage.