Anticipating more stringent maximum contaminant level controls, the Southern Nevada Water System initiated a pilot-scale study of the use of chlorine dioxide to reduce total trihalomethane concentrations in its distributed water. A preliminary investigation revealed that constructing and operating a granular activated carbon postreactor would cost between $60 and $100 million. Comparing the cost and effectiveness of chlorine dioxide with that of GAC showed that Cl02 would be the better alternative for the system. Design considerations and analytical techniques are discussed. Includes 5 references, figures.