Functional connectivity (FC), as assessed with resting-state fMRI, has gained a preponderant position in brain research to study both the healthy and the disordered brain. Yet, FC is a complex metric and despite its widespread use, remains poorly understood. Comparison with structural connectivity assessed with diffusion weighted imaging (DWI) provides only an incomplete representation of the structural connectivity, as DWI-based reconstruction is biased towards larger white matter tracts, poorly identifies crossing fibers, and does not discriminate between mono- and poly-synaptic interactions. In the mouse, viral tracers have been used to extensively map the mouse brain’s monosynaptic structural connectivity (Oh et al., 2014). This provides an unambiguous reference for comparing structural connectivity with FC.C57BL/6 mice (n=14) were anesthetized with isoflurane 0.5% and medetomidine 0.05mg/kg bolus and 0.1mg/kg/h infusion as in (Grandjean et al., 2014). Functional imaging took place on a 9.4 Bruker biospec with a linear volume coil for excitation and a 2x2 phased-array receiver cryogenic coil. Multi-echo gradient-echo EPI were acquired with TE=11, 17, 23ms, TR=1500ms, FA=60°, 600 volumes, and an in plane resolution of 0.3mm². ME-EPI were reconstructed and denoised using meica.py script (Kundu et al., 2012) (AFNI, http://afni.nimh.nih.gov/) , and coregistered to the AMBMC mouse template. Viral tracer maps were downloaded from the Allen Brain Institute website (http://connectivity.brain-map.org/). The Allen template was coregistered to the AMBMC template using ANTS (http://picsl.upenn.edu/software/ants/), and the transformation were applied to the tracer maps. Injection coordinates were used as seed for analysis of the resting-state fMRI data. A winner takes all method was used on 20 injection sites originating in the isocortex previously used in (Oh et al., 2014). The strongest connectivity of the 20 injection sites was color-coded for each voxel of the isocortex, striatum, and thalamus. Two spheres for each voxel indicate the first and second strongest connectivity value. Spearman’s correlation was used to compare tracer-based and functional connectivity values from each injection site for each voxel.Comparison from the structural connectivity matrix estimated in Oh et al., 2014 and the corresponding FC matrix estimated from resting-state fMRI reveal high degrees of similarities between the two (Fig. 1). In particular the patterns of connectivity between the two matrices for interactions within the ipsi- and contral-lateral isocortex indicate similar features. Interactions from the ipsi- to contra-lateral striatum are mostly absent in the matrix based on viral tracer, but are strongly represented in the FC matrix. In contrast, thalamus to ipsi-lateral isocortex interactions are diminished in FC, but present in the tracer data. These changes are highlighted in selected cross-sectional maps (Fig. 2). Tracer maps indicate high degree of correspondence with FC maps for injections and corresponding seed in the isocortical ROIs such as the cingulate cortex and sensory cortex. Injections in the striatum indicate no contra-lateral mono-synaptic projection, despite strong FC. Injection in the thalamus reveals projections to the cortex, which were not apparent in the FC map. Winner takes all analysis indicates that the isocortex could be parcellated in a similar fashion using either tracer-based or functional connectivity, leading also to strong correlations at the voxel level between the two modalities. This remained the case for the striatum, but not for the thalamus.MRI-based connectivity analysis has gained an important position in brain research in the past decade. Yet, structural and functional connectivity are difficult to bridge due to the different nature of the parameters assessed, and the indirect underlying measurements. In mice, retrograde tracing techniques using viral tracer have provided a reference tool for such comparisons (Oh et al., 2014). Here, we report an excellent overlap between brain structure and function in the isocortex, indicating that for this brain region FC is well explained by mono-synaptic projections. In sub-cortical regions, the relationship becomes more complex; with FC in the striatum seem to depend on poly-synaptic interactions, likely relayed via the cortex, and thalamic FC being mostly absent. Winner takes all approach further indicates that the mouse isocortex and striatum can be parcellated in a similar way using either tracer data or FC. In conclusion our results highlight resting-state in the mouse as a highly attractive method to investigate the brain function, due to its close correspondence with the structural basis, and its potential to elucidate subtle changes in connectivity strength non-invasively and in non-terminal experiments, allowing to follow animals over long period of time, such as during the development of a pathology or pharmacological intervention.