Aims: How mitochondrial reactive oxygen species (ROS) impact physiological function may depend on the quantity of ROS generated or removed, and the subcellular microdomain in which this occurs. However, pharmacological tools currently available to alter ROS production in vivo lack precise spatial and temporal control. Results: We used CRISPR/Cas9 to fuse the light-sensitive ROS-generating protein, SuperNova to the C-terminus of mitochondrial complex II succinate dehydrogenase subunits B (SDHB-1::SuperNova) and C (SDHC-1::SuperNova) in Caenorhabditis elegans to localize SuperNova to the matrix-side of the inner mitochondrial membrane, and to the intermembrane space (IMS), respectively. The presence of the SuperNova protein did not impact complex II activity, mitochondrial respiration, or C. elegans development rate under dark conditions. ROS production by SuperNova protein in vitro in the form of superoxide (O2˙-) was both specific and proportional to total light irradiance in the 540-590 nm spectra, and was unaffected by varying the buffer pH to resemble the mitochondrial matrix or IMS environments. We then determined using SuperNova whether stoichiometric ROS generation in the mitochondrial matrix or IMS had distinct effects on redox signaling in vivo. Phosphorylation of PMK-1 (a p38 MAPK homolog) and transcriptional activity of SKN-1 (an Nrf2 homolog) were each dependent on both the site and duration of ROS production, with matrix-generated ROS having more prominent effects. Furthermore, matrix- but not IMS-generated ROS attenuated susceptibility to simulated ischemia reperfusion injury in C. elegans. Innovation and Conclusion: Overall, these data demonstrate that the physiological output of ROS depends on the microdomain in which it is produced. Antioxid. Redox Signal. 31, 594-607.