We compare connectivity estimates from macaque tract tracing, DWI, resting-state fMRI, and cortical thickness covariance.

A generally poor cross-modal correspondence, evaluated by binarized or weighted comparisons, or through clustering analyses.

DWI and fMRI, along with DWI and CoCoMac, had the best agreement among bimodal comparisons.

Higher correspondence for short, ipsilateral, and homotopic connections, and lower for long and heterotopic connections.

Improved correspondence between human, but not macaque, connectivity estimates, for higher resolution neuroimaging data.