The brain’s physiology may fundamentally limit the achievable spatial and temporal specificity of gradient-echo fMRI. Even if the physiology does not pose such a limitation a better understanding would allow for data analysis techniques that improve the spatial specificity. Microscopy allows for highly detailed investigations of local physiological mechanism and provides a growing knowledge from which fMRI may benefit profoundly. A current challenge is the transition from focal mechanisms to their consequence on the mesoscopic scale of BOLD examinations. In this abstract we present our recent work on this transition using simulations of the BOLD effect.