OBJECTIVE To determine the effects of trazodone treatment on behavioral signs of stress in hospitalized dogs. DESIGN Prospective observational study. ANIMALS 120 client-owned dogs. PROCEDURES Hospitalized dogs administered trazodone (n = 60) were observed for stress-related signs or behaviors ≤ 45 minutes after the drug was administered (time 1) and approximately 90 minutes later (time 2). Dogs that did not receive trazodone (n = 60) were selected to serve as controls for environmental stimuli that could affect behavior and were observed at the same times. Signs or behaviors (scored as present or absent) were assessed individually and grouped into behavioral summation categories (frenetic [lip licking, pacing, panting, spinning, trembling, wet dog shake, whining, and yawning], freeze [averting gaze, pinning back ears, and whale eye sign], or fractious [growling, lunging, showing teeth, and snapping], with lifting of a forelimb and pupil dilation included in all categories). Results were compared between groups and within groups over time. Logistic regression was performed to assess associations between reduction in stress-related signs or behaviors and trazodone administration while controlling for environmental influences. RESULTS Lip licking, panting, and whining were reduced (defined as present at time 1 and absent at time 2) in trazodone-treated but not environmentally matched dogs. The median number of stress-related behaviors and of frenetic and freeze behaviors was significantly lower at time 2, compared with time 1, in trazodone-treated dogs. Odds of reduced panting and reduced frenetic behaviors at time 2 for trazodone-treated dogs were > 2 times those for environmentally matched dogs. CONCLUSIONS AND CLINICAL RELEVANCE Results indicated that trazodone administration reduced stress-related signs and behaviors in hospitalized dogs and may thereby improve patient welfare.