The unique access possible within a mine was utilized to detonate controlled explosions throughout the area monitored with a 3D underground seismic network. Measured P-wave travel times were inverted to compute a controlled source velocity image. The arrival times from these controlled sources were also used as simulated passive sources and inverted for a velocity image, which was compared to the controlled source image. Two passive source imaging techniques were used, a simultaneous velocity inversion-hypocenter relocation method and an independent hypocenter relocation and velocity inversion method. The simultaneous inversion method provided the best correlation between the controlled source and simulated passive source images. Only slight improvement was found from multiple iterations of the simultaneous inversion, suggesting that the initial hypocenter estimate is sufficiently close to the “true” location for the problem to be essentially linear. The independent hypocenter relocation and velocity inversion method also provided an image that correlated well with the controlled source image. Good correlation was also found in a comparison of the controlled source image with a passive source image calculated with mining-induced microearthquakes. The correlation was improved with a combined inversion of the explosion and microearthquake data, suggesting that the discrepancies may be due to nonuniqueness of inversions of the individual data sets.