The power spectrum and the voxel intensity distribution (VID) are two of the main proposed summary statistics to study line-intensity maps. We reformulate the derivation of the VID in terms of the local overdensities and derive for the first time an analytic covariance between the VID and the line-intensity mapping power spectrum. We study the features of this covariance for different experimental setups and show that we can recover similar results to simulation-based covariances. With this formalism, we also compute the cosmic variance contribution to the VID uncertainty, which we find to be subdominant with respect to the standard variance from Poisson sampling. Our results allow for general joint analyses of the VID and the line-intensity mapping power spectrum.