To assess and improve existing monitoring protocols for sharp‐tailed grouse (Tympanuchus phasianellus) in the eastern Upper Peninsula of Michigan, we used data from 58 radio‐collared grouse (46 M, 12 F) monitored within 3 openland landscape types: a xeric, conifer‐dominated site, a wetland‐dominated site, and a site dominated by low‐intensity agriculture. We used lek counts and radio telemetry to determine lek attendance rates, factors affecting lek attendance rates, lek fidelity, and inter‐sexual variation in these parameters. Our analysis indicated lek attendance varied with respect to sex of bird, day of year, time after sunrise, and wind speed. Peak male lek attendance rates exceeded those of females by up to 40%, and peak lekking activity for both males and females occurred during the second and third weeks of April. Male lekking activity occurred earlier and was sustained longer than that of females. Lekking activity was negatively related to time of day and wind speed. We observed strong lek fidelity as radio‐collared birds attended a primary lek 94% of the time, indicating a low probability of multiple counting of individual birds. We also proposed a method to adjust lek count data for the probability that birds are on a lek during lek counts. Our proposed method can be used by researchers and managers to improve estimates of the number of birds attending a lek by reducing the negative bias associated with observed counts. © 2011 The Wildlife Society.