Resting state functional connectivity (FC), which measures correlation of spontaneous hemodynamic signal (HemoS) between brain-areas, is becoming an important noninvasive tool to study brain networks. Recent studies showed overall similarity between spatial patterns of HemoS-based FC and underlying neuronal activity patterns (Nir et al., 2008; Scholvinck et al., 2010; Wang et al., 2012; Wang et al., 2013). However, since these studies compared neuronal activity and HemoS only at selected regions, it remains unclear to what extent HemoS reflects precise spatiotemporal patterns of underlying spontaneous neuronal activity in the whole brain. To address this question, we simultaneously recorded HemoS and neuronal activity based on calcium signal (CaS) in transgenic mice expressing genetically encoded calcium indicator (GCaMP3) in excitatory neurons (Zariwara et al., 2012). HemoS was measured at 5 Hz by illuminating the brain with red light (625 nm) and collecting the reflected light through a red filter (600 nm long-pass). CaS was measured at 10 Hz by exciting GCaMP3 with blue light (470 nm) and collecting the epifluorescence through a green filter (520/35-25 nm). Consistent with the presence of hemodynamic delay, HemoS lagged behind simultaneously recorded CaS by about 2 sec. Furthermore spatial patterns of CaS-based FC and HemoS-based FC were highly correlated, suggesting overall similarity between spatial correlation patterns of spontaneous HemoS and neuronal activity. However, closer examinations revealed that coupling between spatiotemporal patterns of CaS and HemoS was not perfect. Most strikingly, transient periods of global brain activity frequently observed in CaS were sometimes not detected in simultaneously recorded HemoS. Similarly, snapshots of CaS that resembled FC-maps were not always reflected in simultaneously recorded HemoS.