Understanding effects of location and timing of harvest seasons on mortality of ducks and geese from hunting is important in forming regulations that sustain viable waterfowl populations throughout their range. During 1990 and 1991 we alternately marked 80 hatching year (HY), female mallards along the Vermont–Quebec border; half with radio-transmitters and bands and half with only aluminum leg bands. We monitored radio-marked ducks daily and recorded survival status weekly for 15 weeks from August until December each year. Mallard mortalities began 25 September when the hunting season opened in the Province of Quebec, Canada. Overall survival of mallards at week 10 did not differ between years (0.51 in 1990 vs. 0.43 in 1991) or differ from that of HY American black ducks (0.44 females, 0.42 males) based on proportional hazard analysis in a Bayesian framework. The mortality rates for mallards from hunting (0.47) and causes unrelated to hunting (0.06) were similar between years and to those rates for HY black ducks at that same site. Hunter harvest accounted for most of the mortality recorded during this study and illegal feeding (i.e., baiting) at sites just before and during the hunting season was observed. Females with lower body condition index had greater mortality rates; a 1-standard-deviation increase in condition index would reduce mortality hazard by about 29%. Management options that may increase mallard survival in the area include lowering daily bag limit in Quebec and suspending split hunting seasons in Vermont that increase harvest, delaying opening date of hunting in Quebec to allow for increased body condition before hunting season opens, and improving enforcement of baiting restrictions.