The green-winged teal (Anas crecca carolinensis) is an important harvest species, yet we know relatively little about its population ecology. We investigated aspects of green-winged teal population ecology of potential importance to waterfowl managers. We used recoveries of green-winged teal banded during winter (1950-89) to establish banding reference areas and estimate survival and band recovery rates. We used cluster analysis based on similarities in recovery patterns to group banding degree blocks into 8 minor and 5 major reference areas describing the principal wintering range of green-winged teal in North America. We then estimated survival and recovery rates of green-winged teal banded in these areas. Mean annual survival rate estimates across years and reference areas were similar (P gt 0.05) for males (0.55, cxa SE = 0.022) and females (0.51, cxa SE = 0.057). Mean annual recovery rate estimates were larger for males (0.033, cxa SE = 0.0017) than for females (0.024, cxa SE = 0.0024) (P lt 0.01). There was little evidence of temporal variation in survival or recovery rates for most datasets. There was evidence of geographic variation in survival rates among major reference areas for males (P = 0.04) but not for females (P = 0.30). We recommend that analyses be conducted on greenwinged teal banded during preseason to further investigate possible sex specificity of survival rates and to address questions about the relationship between harvest rates and survival.