Using a combination of modeling and online water quality instrumentation, utilities can produce real-time predictions of trihalomethane (THM) formation in their distribution system. In this study, an empirical model was used to predict THM formation and embedded into a full-scale supervisory control and data acquisition (SCADA) system. Online water quality instrumentation provided input values to the THM model for total organic carbon (TOC), pH, and temperature. A hydraulic model was also embedded in SCADA, and provided residence time input values for the wholesale portion of the distribution system, which included large lateral pipelines, reservoirs, and pump forebays. Results from a 3-year evaluation showed that predicted THM concentrations were within 10 μg/L of the measured value 81% of the time. Real-time model predictions can provide an effective way to monitor the formation of THMs in a full-scale distribution system.