The golden-winged warbler (Vermivora chrysoptera) is a Neotropical migratory songbird listed as a “Bird of Conservation Concern” by the U.S. Fish and Wildlife Service. To manage golden-winged warblers, it is important to develop effective survey techniques for conservation research and monitoring. We conducted point counts in 1–8-year-old aspen (Populus sp.) stands in the northern Lower Peninsula of Michigan (USA), during 2011, to estimate detection probability of golden-winged warblers, with and without an electronic broadcast of a golden-winged warbler song (i.e., audio lure). We compared audio lure effectiveness for detecting golden-winged warblers during fixed- (50-m radius) and variable-radius point counts. Golden-winged warbler detection estimates were = 0.84 (95% CI = 0.39–0.98) and = 0.22 (0.11–0.40) for fixed-radius point counts, with and without audio lure, respectively. For variable-radius point counts, golden-winged warbler detection estimates were = 0.79 (CI = 0.51–0.93) and = 0.57 (0.42–0.71), with and without audio lure, respectively. We also estimated the number of 3-minute sub-counts required to achieve detection probability ≥95% for both radii, with and without audio lures. We found that 2 sub-counts with audio lure resulted in >95% detection probability for golden-winged warblers at both radii. Without audio lure, 12 and 4 sub-counts were required for fixed and variable radii, respectively. Regardless of survey technique, golden-winged warbler detection probability was always <1.0, which highlights the importance of accounting for imperfect detection of golden-winged warblers. Our results indicate that the use of an audio lure is an effective design technique for improving detectability of golden-winged warblers. © 2014 The Wildlife Society.