We examined the vulnerability of mallards (Anas platyrhynchos) to hunting with a spinning-wing decoy (SWD) during the 2001 and 2002 autumn hunting seasons in Manitoba, Canada. Between 8 September and 11 November 2001, we conducted 72 experimental marsh hunts, and between 8 September and 30 October 2002, we conducted 27 experimental marsh and 55 experimental field hunts. Each hunt consisted of a series of equal and alternating 15-minute experimental (SWD on) and control (SWD off) periods, separated by a 3-minute buffer. Duration of total hunts ranged from 1.0–3.0 hours, with an average of 1.4 ± 0.5 hours. Experimental marsh hunts in 2001 and 2002 indicated that mallards were 1.9 times more likely to fly within gun range (P < 0.001), the mean number killed/hour/hunter was 5.0 times greater (P < 0.001), and the crippling rate was 1.7 times lower when the SWD was on than off (P = 0.02). Additionally, body condition index (size-adjusted body mass) of harvested mallards was greater when the SWD was on than off (P = 0.03). Experimental field hunts in 2002 indicated that mallards were 6.4 times more likely to fly within gun range (P < 0.001), the mean number killed/hour/hunter was 33 times greater (P < 0.001), and, although similar in the early season (before 5 October), crippling rate was 3.7 times lower when the SWD was on than off during the late season (P = 0.047). A SWD activity by age interaction (P = 0.05) indicated that when the SWD was on, harvested adult mallards had a greater body condition index than juveniles but were similar when the SWD was off. Use of a SWD increased mallard flock response, provided more shooting opportunity to hunters, increased kill rate, reduced crippling rates, and, in marsh hunts, resulted in mallards harvested with a greater body condition index. Lastly, quantifying SWD use throughout Canada and the United States will be necessary to estimate its impact on overall harvest.