With 6 subspecies described, the rough-footed mud turtle (Kinosternon hirtipes) has been considered a species with high morphological diversity. One subspecies, K. h. megacephalum, is already extinct. The remaining subspecies are poorly studied, with very limited information available only for K. h. murrayi (a widespread subspecies). The remaining taxa (K. h. tarascense, K. h. chapalense, K. h. magdalense, and K. h. hirtipes) are microendemic and restricted to endorheic valleys in the Mexican Transvolcanic Belt. Using baited fyke nets, hoop traps, and a seine, we surveyed for these microendemic subspecies for 3 yrs across their known distributions. With the data gathered in the field and published information, we conducted a population viability analysis (PVA) to model the minimum characteristics needed to improve population growth under 3 scenarios (optimistic, intermediate, and pessimistic). Very few K. hirtipes turtles were collected. No K. h. chapalense were located in Lake Chapala, but we did capture 4 individuals in Lake Zapotlan. Similarly, only 6 individuals of K. h. magdalense were captured. No K. h. hirtipes individuals were collected in the Valley of México. Two populations of K. h. tarascense were located in the Pátzcuaro basin: the first populations reported for that subspecies. The PVA showed that harvest on adults in populations with fewer than 200 turtles significantly compromises population persistence. However, a population size of 200 and limited or no harvest are enough to increase population size and persistence for at least 100 yrs. Assurance colonies and head-starting may be the only chance for long-term survival of some of these microendemic turtle lineages.