Understanding the nature of the link between neuronal activity and BOLD signal plays a crucial role i) for improving the interpretability of BOLD images and ii) on the design of more realistic models for the integration of EEG and fMRI. The aim of this study is to investigate the neural mechanism underlying hemodynamic behavior in a series of visual stimulation frequencies and explore possible implications for the neurovascular coupling. We studied the relationship between electrophysiological and hemodynamic measures by performing simultaneous steady state electroencephalography (EEG) and fMRI recordings in a healthy human subject during a series of visual stimulation frequencies (6 Hz, 8 Hz, 10 Hz, 12 Hz). BOLD amplitudes were computed for voxels within an anatomical mask which was obtained by mapping the significantly active voxels using general linear modelling (GLM) on fMRI data. On the same anatomical map, EEG power time series belonging to the fundamental frequency and its harmonics due to the stimulation are estimated using a distributed source imaging technique. The neuronal efficacies which represent the vascular inputs driving the BOLD response are estimated by use of an extended version of Balloon model. A nonlinear relationship is demonstrated between the mean EEG source powers and the neuronal efficacies driving the BOLD response. The result suggests that BOLD signal which is an indicator of the metabolic demand of both synchronized and non-synchronized neuronal activities; changes independent of EEG activity which is a measure sensitive to the synchronicity of neuronal activity.