Using 20 years of demographic and genetic data from four populations of banner-tailed kangaroo rats (Dipodomys spectabilis), we asked whether dispersing individuals gain benefits during adulthood that might compensate for the substantial survival costs they experience as juveniles. Compared to philopatric animals, within- and between-population dispersers gained no measureable advantages in adult survival, fecundity, or probability of recruiting offspring to adulthood. Restricting analyses to members of two central populations living more than 15 times the median dispersal distance from the study site edge, and using peripheral populations only to detect dispersal or offspring recruitment "offsite," did not change this result. Population density during year of birth had small negative effects on adult survival and fecundity, but there were no interactions with dispersal phenotype. We found no evidence that dispersers gained access to superior habitat or that their offspring suffered less inbreeding depression. Our results are consistent with fitness advantages being indirect; by leaving, dispersers release their kin from competition. Our results are also consistent with the possibility that dispersal is the "best of a bad lot." If dispersal is a conditional strategy, then the benefits may be obscured in observational data that compare dispersing individuals to philopatric individuals rather than to individuals whose dispersal phenotype is experimentally manipulated.