Sunlight-mediated inactivation of microorganisms is a low-cost approach to disinfect drinking water and wastewater. The reactions involved are affected by a wide range of factors, and a lack of knowledge about their relative importance make it challenging to optimize treatment systems. To characterize the relative importance of environmental conditions, photoreactivity, water quality, and engineering design in the sunlight inactivation of viruses, we modeled the inactivation of three – human adenovirus and two bacteriophages – MS2 and phiX174 – in surface waters and waste stabilization ponds by integrating solar irradiance and aquatic photochemistry models under uncertainty. Through global sensitivity analyses, we quantitatively apportioned the variability of predicted sunlight inactivation rate constants to different factors. Most variance was associated with the variability in and interactions among time, location, non-purgeable organic carbon (NPOC) concentration, and pond depth. Photolysis quantum yield of the virus outweighed seasonal solar motion in the impact on inactivation rates. Further, comparison of simulated sunlight inactivation efficacy in maturation ponds under different design decisions showed reducing pond depth can increase the log inactivation at the cost of larger land area, but increasing hydraulic retention time by adding ponds-in-series yielded greater improvements in inactivation.