The spatiotemporal and longitudinal monitoring of cellular processes occurring in tumors is critical for oncological research. We focused on glioblastoma multiforme (GBM), an untreatable highly vascularized brain tumor whose progression is thought to critically depend on the oxygen and metabolites supplied by blood vessels. We optimized protocols for orthotopic GBM grafting in mice that were able to recapitulate the biophysical constraints normally governing tumor progression and were suitable for intravital multiphoton microscopy. We repeatedly imaged tumor cells and blood vessels during GBM development. We established methods for quantitative correlative analyses of dynamic imaging data over wide fields in order to cover the entire tumor. We searched whether correlations existed between blood vessel density, tumor cell density and proliferation in control tumors. Extensive vascular remodeling and the formation of new vessels accompanied U87 tumor cell growth, but no strong correlation was found between local cell density and the extent of local blood vessel density irrespective of the tumor area or time points. The technique moreover proves useful for comparative analysis of mice subjected either to Bevacizumab anti-angiogenic treatment that targets VEGF or to AMD3100, an antagonist of CXCR4 receptor. Bevacizumab treatment massively reduced tumoral vessel densities but only transiently reduced U87 tumor growth rate. Again, there was no correlation between local blood vessel density and local cell density. Moreover, Bev applied only prior to tumor implantation inhibited tumor growth to the same extent as post-grafting treatment. AMD3100 achieved a potent inhibition of tumor growth without significant reduction in blood vessel density. These results indicate that in the brain, in this model, tumor growth can be sustained without an increase in blood vessel density and suggest that GBM growth is rather governed by stromal properties.