Movements and behavior of 89 young mallards (Anas platyrhynchos) and 48 young wood ducks (Aix sponsa) were monitored on a 932-km2 study area in north-central Minnesota in late summer and fall, 1972-74, with telemetry, visual observation, and aerial surveys. Initial flights of both species were confined to the natal (brood) marsh; first flights away from the natal marsh occurred in the third week after fledging in both species. First flights of young mallards and wood ducks away from their natal marshes were not significantly different between the sexes (mallard, mean = 4.95 km for females and 5.83 km for males; wood ducks, mean = 2.31 km for females and 2.64 km for males). However, flights away from the brood marshes by wood ducks were significantly shorter than for mallards.As young mallards and wood ducks grew, their daytime use of the natal marshes decreased in an irregular pattern as both species began daily flights between day- and night-use areas. Locally reared mallards made longer daily flights between use areas than did wood ducks, but wood ducks changed use areas with greater frequency before 1 October. Despite often extensive movements, most locally reared mallards and wood ducks remained in the vicinity of their brood marshes throughout fall until migration.Movement of young birds to new habitat was not the result of random searching and thus fortuitous discovery of nearby areas. Instead, birds seemed to learn of new habitat and develop movement patterns by associating with other birds; locally reared young always moved in the company of flocks of conspecifics that included adults and older immatures.Differences in movement patterns between the sexes of young birds and between young and adult birds cause them to be differentially distributed by age and sex on and near the breeding grounds. These differences are ultimately reflected in the distribution of the hunter harvest. We have interpreted generalizations about such phenomena, developed from analysis of continent-wide mallard banding data, using our data obtained from individually marked birds. We document (a) greater distances moved by early than by late-hatched young in the postbreeding period before migration, (b) differential movement of age and sex cohorts that explains greater hunting mortality of young than adults and of females than males near natal marshes, (c) differences in length and timing of postbreeding movements of adult male mallards and the postfledging movements of immature male mallards that help explain the northerly continental recovery distribution of young males, and (d) differential timing and rate of movement by birds through harvest areas (early departure of males and some return of females to natal marshes after the beginning of hunting) that explain differences in the timing of hunting season recoveries. Behavioral differences between the age and sex cohorts in the fall waterfowl population on and near their breeding grounds in north-central Minnesota can explain observed differences in survival and recovery rates of adult and young birds.The behaviors observed suggest to us that restrictive harvest regulations such as small-area closure may have little or no local benefits at the breeding grounds because premigratory assemblages of birds make extensive movements. In particular, protection of postbreeding adult females and locally reared young might only occur by closing large areas or scheduling extreme delays in the season opening, neither of which may be compatible with equitably apportioning waterfowl harvest at higher latitudes. Additional research on the local effects of restrictive regulations, and on age- and sex-specific differences in the timing, rate, and direction of fall movements of postbreeding waterfowl is needed.