We examined how hunter behavior, environmental covariates, and mallard (Anas platyrhynchos) and grey duck (A. superciliosa) population indices affected per capita harvest, hunter effort (i.e., hours hunted), and hunter participation (i.e., license sales) between 1997 and 2012 in the Eastern Fish and Game Region of New Zealand. Additionally, we examined how total annual hunter effort and harvest affected annual survival and harvest rates (i.e., the proportion of the population that is harvested). Per capita harvest increased with hunter effort and bag limits; hunter effort decreased over time, but effort and participation increased with mallard population size. Juvenile harvest rates were greater than for adults and negatively associated with population size. The relationship between harvest regulations and harvest rates was inconsistent. The 44‐day seasons had greater juvenile harvest rates than the 57‐ and 72‐day seasons. Similarly, years with a 7‐bag limit had higher juvenile harvest rates than years with a 10‐bag limit. Hunter effort affected annual survival rates, especially for females. Alternatively or concordantly, hunter effort may be a surrogate for population size and thus, survival rate may have been density dependent. The relationship between harvest and density‐dependent mortality may in part be augmented by hunter behavior; fewer hunters hunted for fewer hours in years with relatively few birds. Our results suggest bag limits are more effective than season length at managing harvest; reducing bag limits to <2 birds/day from ≥7 could decrease harvest by as much as 50%. Furthermore, regulation consistency, better education, and enforcement of season regulations may improve harvest management; 58% of active hunters reported they shot mallards or grey ducks after the close of the 31‐day season, which accounted for 13% of total harvest.