Objective: To identify modifiable risk factors for depression in people with spinal cord injury (SCI). Design: Cross-sectional survey. Setting: Outpatient and community settings. Participants: Community-residing people with SCI (N=244; 77% men, 61% white; mean age, 43.1y; 43% with tetraplegia) who were at least 1 month postinjury. Interventions: Not applicable. Main Outcome Measures: Depression severity (Patient Health Questionnaire-9 [PHQ-9]), physical activity (International Physical Activity Questionnaire [IPAQ]), pleasant and rewarding activities (Environment Rewards Observation Scale [EROS]), and self-efficacy to manage the effects of SCI (Modified Lorig Chronic Disease Self-Management Scale). Results: Greater depression severity was associated with being 20 to 29 years of age, not completing high school, not working or attending school, and being ≤4 years post-SCI. After controlling for demographic and injury characteristics (adjusted R 2=.13), lower EROS scores (change in adjusted R 2=.34) and lower self-efficacy (change in R 2=.13) were independent predictors of higher PHQ-9 scores. Contrary to predictions, physical activity as measured by the IPAQ did not predict depression severity. Conclusions: Our findings suggest that having fewer rewarding activities, and to a lesser extent, having less confidence in one's ability to manage the effects of SCI are independent predictors of greater depression severity after SCI. Interventions such as behavior activation, designed to increase rewarding activities, may represent an especially promising approach to treating depression in this population. © 2012 American Congress of Rehabilitation Medicine.