Wolf (Canis lupus) kill rates are fundamental to understanding predation, but are not well known at low moose (Alces alces) densities. We investigated kill rates of 6 wolf packs (2–10 wolves/pack) during 2 winters on the Yukon Flats, a region of eastern Interior Alaska where moose were the sole ungulate prey of wolves occurring at densities <0.2 moose/km2. Our objectives were to compare kill rates with those from areas of greater moose densities, and to determine potential trends in kill rates across the winter. We located moose killed by wolves in February–March 2009, and November 2009–March 2010 using aerial tracking techniques and global positioning system (GPS) location clusters. Wolves killed more moose in early than late winter (βMONTH = −0.02 moose/pack/day, 95% CI = −0.01 to −0.04), and kill rate estimates (mean, 95% CI) were greatest in November (0.033 moose/wolf/day, 0.011–0.055) and least in February (0.011, 0.002–0.02). Kill rates were similar between February and March 2009 (0.019 moose/wolf/day, 0.01–0.03) and 2010 (0.018, 0.01–0.03). Prey composition was primarily adult females (39%) and young-of-the-year (35%). We attribute an elevated kill rate in early winter to predation on more vulnerable young-of-the-year. Kill rates in our study were similar to those from other studies where moose occurred at greater densities. We suggest that very few, if any, wolf–moose systems in Alaska and the Yukon experience a density-dependent phase in the functional response, and instead wolves respond numerically to changes in moose density or availability in the absence of alternative prey. Through a numerical response, wolf predation rates may approximate the annual growth potential of the moose population, contributing to persistent low densities of moose and wolves on the Yukon Flats. Published 2013. This article is a U.S. Government work and is in the public domain in the USA.