This project evaluated a medium-pressure ultraviolet (UV) reactor (treating 3 million gallons per day [mgd]) downstream from a conventional surface water treatment plant to identify process design and operations issues. The results showed that adequatedisinfection could be achieved, but monitoring disinfection effectiveness was difficult. Across the water quality range studied, >3 log10 inactivation of heterotrophic bacteria was consistently achieved. During biodosimetry challenges, >2 log10 inactivation of MS-2 coliphage was accomplished at 75 percent of the reactor's lamp capacity. However, very small increases in turbidity--from 0.05 to 0.10 ntu--caused sensor readings to drop 34 percent. This response calls into question the ability of existing sensors to reliably monitor UV performance. UV sensors were checked against a radiometer that directly measured UV irradiance, a procedure that may improve monitoring procedures for large-scale UV reactors. To make UV dose measurement more operator-friendly, a further characterization of water quality effects on sensor performance is necessary. Includes 11 references, tables, figures.