Accurate assessments of local population size of the black-tailed prairie dog (Cynomys ludovicianus) are essential because of their overall decline and importance to prairie ecosystems. We describe the use of mark–resight methodology to estimate black-tailed prairie dog population size and density. Study colonies include isolated urban habitat fragments in Denver, Colorado, USA, and unfragmented control colonies in the Pawnee National Grassland, USA. We compare results from various mark–resight estimators to those derived from linear transformations of visual counts of active prairie dogs. Our results suggest that mark–resight methods are feasible in both urban and rural systems, and reveal extremely high densities for isolated prairie dogs in urban sites. Our methodology can be used to obtain reliable, unbiased estimates of local population size and density.