The vascular space occupancy (VASO) fMRI method probes changes in cerebral blood volume (CBV) under various physiological states, including neuronal activation in humans. However, it requires a careful choice of sequence parameters because the blood oxygen-level dependent (BOLD) effect offsets the VASO signal. Assessing this BOLD contamination as a function of pulse sequence parameters would improve the quantification of CBV changes with VASO. However, this task requires knowledge of the cerebral vascular geometry of the MRI voxel. Towards this end, optical microscopy can provide high-resolution 3D images of vasculature. Here, we use detailed angiograms of rodent brain acquired with two-photon microscopy to model fMRI signals (VASO and BOLD) from first principles using Monte Carlo diffusion of water protons. We present quantitative plots of VASO together with intra- and extravascular BOLD fractional signal changes as a function of echo time (TE), for spin echo (SE) and gradient echo (GRE) pulse sequences, at low to ultra-high magnetic fields. Our results indicate that at 3T, the BOLD contamination of the VASO response is under 12% for GRE and 2% for SE up to TE=6 ms, but this contamination is significantly higher at 7T and above. We also found GRE BOLD intravascular contributions of 85% at 1.5T, 55% at 3T and 4% at 7T and SE intravascular contributions of 70% at 1.5T, 40% at 3T and 10% at 7T. These results may provide important information to optimize the pulse sequence timing in human VASO and BOLD fMRI, leading the way to a wider application of these fMRI techniques in healthy and diseased brain.