Assessment of vascular properties is essential to diagnosis and follow-up and basic understanding of pathogenesis in brain tumors. In this study, a procedure is presented that allows concurrent estimation of cerebral perfusion, blood volume, and blood-brain permeability from dynamic T(1)-weighted imaging of a bolus of a paramagnetic contrast agent passing through the brain. The methods are applied in patients with brain tumors and in healthy subjects. Perfusion was estimated by model-free deconvolution using Tikhonov's method (gray matter/white matter/tumor: 72 +/- 16/30 +/- 8/56 +/- 45 mL/100 g/min); blood volume (6 +/- 2/4 +/- 1/7 +/- 6 mL/100 g) and permeability (0.9 +/- 0.4/0.8 +/- 0.3/3 +/- 5 mL/100 g/min) were estimated by using Patlak's method and a two-compartment model. A corroboration of these results was achieved by using model simulation. In addition, it was possible to generate maps on a pixel-by-pixel basis of cerebral perfusion, cerebral blood volume, and blood-brain barrier permeability.