Pumas (Puma concolor) are the primary predator of adult mule deer (Odocoileus hemionus) throughout most of arid New Mexico, and predation by pumas is popularly believed to limit mule deer populations. Predation can limit population size of prey if it is primarily additive or acts in addition to other mortality factors. Thus, clarifying the impact of predation is important because of misconceptions regarding the act of predation (i.e., killing an individual) rather than the effect (i.e., effects on population demography) cloud discussions of the issue. Herein, we tested the population-level question of whether puma predation on adult female mule deer is compensatory or additive mortality in multiple mule deer populations in New Mexico. We modeled annual survival rate as a function of cause-specific mortality rate. For this conservative test, a slope = 0 indicated complete compensation, whereas a slope = –1 indicated complete additivity. For all populations, the corrected slope of predation ultimately attributable to pumas was 0.00, ultimate and proximate puma predation combined was 0.10, ultimate puma predation and unknown causes of mortality combined was –0.28, and ultimate and proximate puma predation and unknown deaths combined was –0.23. Thus, puma predation appeared primarily compensatory. Despite small sample sizes, individual populations showed similar patterns indicating the compensatory nature of puma predation. Primarily compensatory predation for mule deer reflected the relatively low condition of deer. Survival was most strongly tied to individual condition in these populations, and a priori condition of puma-predated mule deer was lower than the population means. Compensatory mortality requires predisposition whether caused by decreased condition, high population density, or other factors. If conditions that predispose individual deer to mortality are present, managers should not assume that predation is limiting, regardless of predation rates.