Population structure of Harlequin Ducks (Histrionicus histrionicus) wintering in the Strait of Georgia, British Columbia, was evaluated by generating age, sex, paired status and distance-specific movement rates with multi-stratum mark-recapture analyses, and age and sex-specific movement distances through surveys of marked individuals. Annual movement distances and rates did not differ by sex, but only 2–4% of adults (third year and afterthird year) compared to 7–11% of subadults (hatch year and second year) moved among locations per year and distance moved decreased with age. Adults were highly site faithful regardless of sex and paired status. The stepping stone gene flow model estimated the among population component of genetic variance (F_ST) at 0.005, suggesting that winter movement by subadults was sufficient to explain results of previous genetic analyses that detected no fine scale genetic structuring. Seasonal movement rates indicated that at least 95% of individuals molt and winter in the same location, and that annual aggregation at Pacific Herring (Clupea pallasi) spawning sites facilitates demographic mixing and gene flow. Low annual movement rates (0.001) between the northern and southern Strait of Georgia and dispersal by both sexes suggest that a metapopulation distribution may function within the Pacific Coast range, which is relevant to the geographic scale of management. Movement rates and distances suggest that subadult survival rates are particularly vulnerable to underestimation.