Annual waterfowl harvest in North America is estimated through a collaborative and strategic process, with federal harvest surveys the primary method of estimation. Sea duck hunters participating in federal harvest surveys represent a small proportion of the overall waterfowl hunting population, limiting the utility of harvest estimates for sea ducks. The long-tailed duck (Clangula hyemalis) is one such species. To partially address the paucity of long-tailed duck harvest survey information, we conducted in-person hunter surveys from 1 November through 4 December 2016 at a boat launch in Two Rivers, Wisconsin, USA (Lake Michigan), an important area for long-tailed duck harvest within the state. Hunters were present on 15 of 21 survey days, and we surveyed occupants of 62 individual hunting boats on 127 occasions. Long-tailed ducks were the most common (97%) of the 1,431 sea ducks reported harvested by hunters. Hunter harvest of long-tailed ducks averaged 3.8 (95% CI = 3.4, 4.1; range = 0–6) long-tailed ducks/hunter/day. We used count models to evaluate the effects of environmental variables on hunter participation and harvest of long-tailed ducks. Wave height was the most influential predictor variable for hunter participation; an information criterion-based best model (wave height + temperature) indicated that hunter participation decreased by 91.9% (95% CI = 79.3–97.1%) for each 1-m increase in wave height. Long-tailed duck harvest was positively associated with air temperature; the relationship indicated a 9.5% (95% CI = 6.2–12.9%) increase in long-tailed duck harvest with each degree increase in temperature. Our results contribute to the understanding of waterfowl hunter participation, hunter preferences, and harvest on Lake Michigan and can inform managers as they assess regulatory frameworks for sea duck hunting.