Blood oxygenation level dependent (BOLD) functional magnetic resonance imaging (fMRI) is widely used as a measure of neuronal activity, despite an incomplete understanding of the hemodynamic and neural bases for BOLD signals. Recent work by Lee and colleagues investigated whether activating genetically specified neurons elicits BOLD responses. Integrating optogenetic control of specific cells and fMRI showed that stimulating excitatory neurons triggers a positive BOLD signal with conventional kinetics locally and delayed weaker BOLD signals distally.Blood oxygenation level dependent (BOLD) functional magnetic resonance imaging (fMRI) is used to noninvasively study brain function, primarily in research, although the technique also has clinical applications such as the preoperative detection of eloquent brain areas, regions of the brain that regulate our senses, movement, and speech. BOLD signals reflect hemodynamic changes in flow and blood volume and exploit intravascular magnetic susceptibility to infer oxygenation and the cerebral metabolic rate of oxygen consumption (Logothetis et al. 2001). To illustrate the increasingly prolific use of fMRI in neuroscientific research, a search on PubMed for "functional magnetic resonance imaging" or related abbreviations in the title or abstract returns 1,666 results within the first 5 months of 2010 alone, more than 11 articles per day. However, the technique has not escaped criticism, particularly from neurophysiologists, since the complexities of the hemodynamic response and its relationship to neuronal activity confound the interpretability of fMRI results. Indeed, it remains unclear which types of neuronal activity are capable of eliciting BOLD signals or even which cell populations contribute (Logothetis 2008). If a cerebral microcircuit receives neuromodulatory input, leading to balanced proportional changes in excitatory and inhibitory activity, it is relatively straightforward to predict the effect on hemodynamic responses in the region and resulting BOLD signal change. However, if cortical excitation and inhibition are driven in opposite directions, resulting in net excitation or net inhibition of our hypothetical cerebral microcircuit, the resulting BOLD signal changes are more difficult to predict (Logothetis 2008), since increased inhibitory activity and reduced spiking may increase local cerebral metabolism (Sokoloff et al. 1977). As well as potentially confusing excitation and inhibition, the interpretability of fMRI activation maps is confounded by an incomplete understanding of the relative contributions of different types of activity to the hemodynamic response, such as local field potentials of various frequencies, multiple unit activity, or spiking of individual neurons. Keep in mind that Ͻ3% of the volume of one typical neuroimaging voxel is occupied by vasculature, 97% being neural matter: neurons, synapses, and glia (Logothetis 2008).