We assessed effects of radio collars on survival of Columbian sharp-tailed grouse (Tympanuchus phasianellus columbianus) in western Idaho during 1983-86. Thirty-eight grouse were colorbanded and equipped with radio collars, and 9 were colorbanded only. Four nonradioed grouse were the only marked birds resighted 1 year following capture. Radio collars did not restrict flight but altered the appearance, sound in flight, and perhaps the behavior of grouse. Avian predators caused most of the mortality and probably preyed on radio-collared birds selectively. Timing of mortality coincided with spring and autumn dancing periods, presence of northern goshawks (Accipiter gentilis) in the study area, and seasonal lows in vegetative cover. Despite poor survival of radio-collared grouse, we detected no decline in the population. J. WILDL. MANAGE. 51(2):468-471 The effect of harness-mounted (backpack) radio packages on the behavior and survival of gallinaceous birds has received much recent attention (Herzog 1979, Warner and Etter 1983, Hines and Zwickel 1985). Some authors have reported that radio harnesses have little effect on galliforms, whereas others have documented reduced survival of radio-equipped individuals (see Hines and Zwickel [1985]). Amstrup (1980) designed a poncho-mounted radio collar that eliminated some of the drawbacks of radio harnesses (e.g., restricted mobility) and used it successfully on sharp-tailed grouse. Subsequently, Small and Rusch (1985) found that survival of radio-collared ruffed grouse (Bonasa umbellus) was higher than that of those with radio harnesses. Although Amstrup's radio collar appeared to be an improvement over backpacks, no study has compared survival of radio-collared vs. nonradioed grouse. As part of a study of habitat requirements of Columbian sharp-tailed grouse, we captured and radio-collared adults in early spring and autumn 1983-85 (see Marks [1986]). A smaller sample of birds was colorbanded and released without radios. Here, we discuss effects of radio collars on survival and factors influencing the seasonal timing of mortality of sharp-tailed grouse in western Idaho. Primary funding was provided by the Boise Dist. Bur. Land Manage., with additional support from the Idaho Dep. Fish and Game and the Dep. Biol., Montana State Univ. A. R. Sands, S. N. Mattise, R. L. Eng, R. E. Autenrieth, and L. J. Nelson deserve thanks for their contributions to the study. The comments of F. C. Zwickel, R. J. Small, and J. W. Connelly improved an earlier version of the paper. STUDY AREA AND METHODS The 20-km2 study area is located in western Idaho about 23 km north of Weiser, Washington County. The vegetation is shrub-steppe dominated by big sagebrush (Artemisia tridentata), low sagebrush (A. arbuscula), and bitterbrush (Purshia tridentata). Bitter cherry (Prunus emarginata), common chokecherry (P. virginiana), Saskatoon serviceberry (Amelanchier alnifolia), and snowbrush ceanothus (Ceanothus velutinus) occur in patches on hillsides. Douglas hawthorn (Crataegus douglasii) is the most common riparian shrub. Sharp-tailed grouse have not been hunted in the study area since 1974. From 1983 to 1985, we captured 46 sharptailed grouse (36 males and 10 females) on dancing grounds using funnel traps, mist nets, and drop nets. Grouse were weighed with a Pesola spring balance, and 38 were equipped with solar-powered radio transmitters attached to herculite ponchos (Amstrup 1980). These packages weighed 13.5-14.5 g, or about 1.72.5% of the body weight of grouse at the time of capture. All captured grouse, including 9 without radios, were marked with unique combinations of 4 colored leg bands. A nonradioed grouse captured in 1983 was recaptured and radiocollared in 1984. We recovered both radio-collared and nonradioed grouse and determined cause of death from field sign (Dumke and Pils 1973). Avian predators normally removed the head, plucked