How does targeting in armed conflict affect public opinion? Armed actors choose between targeting militaries and civilians, and further choose whether to target civilians discriminately or indiscriminately. Existing work suggests these choices are based in part on the effects of targeting on public opinion, yet the causal link between these variables has not been clearly established. We conduct a survey experiment in the Donbass region of Ukraine to analyze this relationship. We find that civilian targeting robustly reduces approval of both the government and separatist forces. Second, we find that the effects of discriminate civilian targeting are generally not statistically distinguishable from those of indiscriminate targeting. Finally, we find that our respondents generally preferred a restrained, rather than reciprocal, response from actors in this armed conflict.