Background: Previous officer-involved shooting (OIS) research has focused primarily on antecedents to fatal shootings, with few studies investigating injury more broadly. Our study examined the factors associated with fatal or non-fatal injury to both civilians and officers during OIS incidents, to better understand how harm might be reduced in the most extreme law enforcement scenarios. Methods: Data included 281 officers involved in 177 unique shooting incidents recorded by Dallas Police Department between 2005 and 2015. Bivariate logistic regression and multivariable generalised estimation equation analyses were used to investigate incident characteristics associated with fatal or non-fatal injury to civilians and officers. Results: Civilian injury occurred in 61% and officer injury in 14% of unique OIS incidents. In adjusted models, multiple shooting officers increased the odds of injury to both civilians (adjusted OR (AOR): 3.22, 95% CI 1.39 to 7.50) and officers (AOR 4.73, 95% CI 1.64 to 13.65). Odds of civilian injury were also significantly higher during the daytime and among non-Hispanic white compared with non-Hispanic black and Hispanic/Latina/o civilians, although a majority of OIS incidents (79%) involved non-Hispanic black or Hispanic/Latino/a civilians. Odds of officer injury were significantly higher for detectives compared with patrol (AOR=9.32, 95% CI 1.85 to 47.03) and during off-duty versus on-duty shootings (AOR=5.23, 95% CI 1.37 to 19.99). Conclusions: Both civilians and officers are at risk for injury during OIS incidents, though to different degrees and with unique risk factors. Additional research is needed to understand whether these results are replicated elsewhere and to further understand the mechanisms of injury.