Onsite wastewater treatment systems (OWTSs or septic systems), when properly sited, designed, operated, and maintained, treat domestic wastewater to reduce impacts on and maintain sustainability of aquatic resources. However, when OWTSs are not performing as expected, they can be a potential source of human fecal pollution to recreational waters, resulting in an increased risk of illness to swimmers. Quantifying the contribution of poor-performing OWTSs relative to other sources of fecal pollution is particularly challenging in wet weather when various sources commingle as they flow downstream. This study aimed to estimate the total load of human fecal pollution from OWTSs in an arid watershed with municipal separate storm sewer systems (MS4). The novel study design sampled HF183, a DNA-based human marker, from six small catchments containing only OWTSs and no other known human fecal sources, such as sanitary sewer collection systems or people experiencing homelessness. Then, the human fecal loading from the representative catchments was extrapolated to the portions of the watershed that were not sampled but contained OWTSs. Flow-weighted mean HF183 concentrations ranged from 104 to 107 gene copies/100 mL across 29 site-events. HF183 mass loading estimates were normalized to the number of parcels per catchment and inches of rainfall per storm event. Assuming the normalized loading estimate was representative, extrapolation to all of the OWTS parcels in the watershed and average annual rainfall quantity illustrated that HF183 loading from OWTSs was a small but measurable fraction of the total HF183 mass loading emanating at the bottom of the watershed. Clearly, other human fecal sources contributed HF183 during storm events in this watershed. The loading estimate approach used in this study could be applied to other watersheds facing similar challenges in prioritizing resources for monitoring and mitigation among co-located human fecal sources.