From 1950 through 1988, the continental breeding population of northern pintails (Anas acuta) varied from 2.0 million to 9.9 million. Because pintails have high fidelity to certain wintering grounds along coasts and large bodies of water, management on these wintering areas may increase population size if changes in winter survival rate are related to changes in population size. I used band-recovery data to estimate survival rates for winter-banded pintails and to test for sex-specific, temporal, and geographic variation in survival rates. Survival rate estimates varied between 0.632 and 0.806 for males, and 0.421 and 0.769 for females. Males had higher (P lt 0.0001) average annual survival rates than females. Limited geographic variation occurred in estimates of average annual survival rates for males, and no variation occurred for females. Males had lower average annual survival rates in the Imperial Valley than in central California (P = 0.007) or in the Gulf Coast (P = 0.092). Little annual variation was found within time periods. However, longer-term variation was found in survival rate estimates for males and females. Males had higher (P = 0.054) average annual survival rates in the Pacific Flyway during 1959-61, a period of drought, breeding-population decline, and restrictive hunting regulations, than during 1950-58, a period with a higher breeding population and liberal regulations. The increase in wintering population size in the Pacific Flyway during the 1970's was associated with a higher average annual survival rate for females in the Pacific Flyway than during the 1950's. Results from the Pacific Flyway suggested that an interaction may exist between population size and the effect of harvest regulations on survival of males. Changes in harvest regulations appeared to have a greater effect at lower population levels.