Gradient and spin echo (GRE and SE, respectively) weighted magnetic resonance images report on neuronal activity via changes in deoxygenated hemoglobin content and cerebral blood volume induced by alterations in neuronal activity. Hence, vasculature plays a critical role in these functional signals. However, how the different blood vessels (e.g. arteries, arterioles, capillaries, venules and veins) quantitatively contribute to the functional MRI (fMRI) signals at each field strength, and consequently, how spatially specific these MRI signals are remain a source of discussion. In this study, we utilize an integrative model of the fMRI signals up to 16.4 T, exploiting the increasing body of published information on relevant physiological parameters. Through simulations, extra- and intravascular functional signal contributions were determined as a function of field strength, echo time (TE) and MRI sequence used. The model predicted previously reported effects, such as feasibility of optimization of SE but not the GRE approach to yield larger micro-vascular compared to macro-vascular weighting. In addition, however, micro-vascular effects were found to peak with increasing magnetic fields even in the SE approach, and further increases in magnetic fields imparted no additional benefits besides beyond the inherent signal-to-noise (SNR) gains. Furthermore, for SE, using a TE larger than the tissue T(2) enhances micro-vasculature signal relatively, though compromising SNR for spatial specificity. In addition, the intravascular SE MRI signals do not fully disappear even at high field strength as arteriolar and capillary contributions persist. The model, and the physiological considerations presented here can also be applied in contrast agent experiments and to other models, such as calibrated BOLD approach and vessel size imaging.