We used immunochemistry, light microscopy and stereological methods for quantitative description of the microvascular network in 13 tissue samples of the human brain. While the tortuosity of microvessels was comparable in all brain parts under study, the length density of microvessels was higher in subcortical grey matter (652.5±162.0 mm-2) and in the cortex (570.9±71.8 mm-2) than in the white matter (152.7±42.0 mm-2). The numerical density of microvessels was higher in subcortical grey matter (3782.0±1602.0 mm-3) and cerebral cortex (3160.0±638.4 mm-3) than in white matter (627.7±318.5 mm-3). We developed simulation software gensei which generates series of images representing three-dimensional models of microvessels with known length density, volume fraction, and surface density. The simulations are statistically similar to real microvessel networks and can be used for computer modelling of brain perfusion.