We assessed mate fidelity, nest-box fidelity, and breeding dispersal distances of American Kestrels (falco sparverius) nesting in boxes in southwestern Idaho from 1990 through 2006. Seventy-seven percent of boxes had different males and 87% had different females where nest-box occupants were identified in consecutive years. High turnover rates were partly a result of box-switching. Forty-eight percent of males and 58% of females that nested within the study area in successive years used different boxes. The probability of changing boxes was unrelated to gender, nesting success in the prior year, or years of nesting experience. Breeding dispersal distances for birds that moved to different boxes averaged 2.2 km for males (max = 22 km) and 3.2 km for females (max = 32 km). Approximately 70% of birds that nested in consecutive years on the study area had a different mate in the second year. Mate fidelity was related to box fidelity but not to prior nesting success or years of nesting experience. Mate changes occurred 32% of the time when the previous mate was known to be alive and nesting in the area. Kestrels that switched mates and boxes did not improve or decrease their subsequent nesting success. Kestrels usually switched to mates with less experience and lower lifetime productivity than their previous mates. The costs of switching boxes and mates were low, and there were no obvious benefits to fidelity. The cost of "waiting" for a previous mate that might have died could be high in species with high annual mortality.