In eastern Oregon, western juniper (Juniperus occidentalis) encroachment has degraded vast amounts of wildlife habitat; thus, the removal of juniper has become a common restoration practice. Vegetative responses following juniper treatment have been well documented; however, wildlife, especially mule deer (Odocoileus hemionus) responses have not. We deployed camera traps at 40 sites within the Phillip W. Schneider Wildlife Area in northeastern Oregon to assess habitat use of wintering mule deer during two winters with different conditions, a mild winter in 2017–2018 and a harsher winter in 2018–2019. We assessed mule deer spatial habitat use using N-mixture models accounting for imperfect detection. We assessed temporal use by fitting kernel density functions to the timestamps of mule deer photographs. Spatial analysis indicated that during a mild winter, mule deer used unburned juniper treatment habitats more than untreated and unburned habitats. Once burned, however, juniper treatment habitats were used the least. During a harsher winter, mule deer spatially responded negatively to total precipitation rather than habitat conditions. Temporal analysis indicated that mule deer did not change their activity pattern between the two winters and used a mosaic of habitats to meet their daily requirements. Our results highlight the importance of juniper removal from encroached sage-steppe habitat but also emphasize the importance of habitat heterogeneity. Our results also demonstrate that wildland fires can negate the benefit of juniper treatment to wintering mule deer.