One tool used for wildlife management is the deployment of radiocollars to gain knowledge of animal populations. Understanding the influence of individual factors (e.g., species, collar characteristics) and landscape characteristics (e.g., forested cover, shrubs, and fencing) on retention of expandable radiocollars for ungulates is important for obtaining empirical data on factors influencing ecology of young-of-the-year ungulates. During 2001–2009, we captured and radiocollared 198 white-tailed deer (Odocoileus virginianus) fawns, 142 pronghorn (Antilocapra americana) fawns, and 73 mule deer (O. hemionus) fawns in South Dakota, Minnesota, and California, USA. We documented 72 (36.4%), 8 (5.6%), and 7 (9.6%) premature (<270 days post-capture) collar losses among white-tailed deer, pronghorn, and mule deer, respectively. Probability of a collar being retained for 270 days was 0.36 (SE = 0.05, 95% CI = 0.27–0.47), 0.91 (SE = 0.03, 95% CI = 0.82–0.96), and 0.87 (SE = 0.05, 95% CI = 0.73–0.94) for white-tailed deer, pronghorn, and mule deer fawns, respectively. Agricultural fencing, which varied among study areas and thus species, seemed to influence collar retention; fencing density was 69% lower in areas where fawns retained collars ( = 1.00 km/km2, SE = 0.1, n = 75) compared with areas where fawns shed collars ( = 3.24 km/km2, SE = 0.1, n = 56) prior to 270 days. Researchers of fawns should consider that radiocollars can be shed prematurely when estimating desired sample size to yield a suitable strength of inference about some natural process of interest.