The goal of this research was to develop a model for predicting free chlorine, Trihalomethanes (THMs), and Haloacetic Acids (HAAs) concentrations in the South Central Connecticut Regional Water Authority (RWA) distribution system. The objective was to develop a model that captures the effects of chlorine dose, temperature and pH without condition-specific calibration; that can handle mixing of waters from different sources with different natural organic matter (NOM) characteristics or applied chlorine doses; and that could be easily modified to model additional byproducts including reactive species. A model for THMs, HAAs and chlorine in water distribution systems was developed and field-tested. The model utilizes a semi-mechanistic representation of chlorine/NOM reaction kinetics. Under this approach, multiple substance concentrations are tracked and local reaction rates are based on the concentrations of the reactants (chlorine and NOM sites). Good agreement between model predictions and field measurements of chlorine residuals, THMs, and HAAs was observed when the model was field-tested in the New Haven, Connecticut, distribution system. The model accounts for chlorine demand due to reactions with corrosion products in unlined iron pipes. There is no other pipe reaction component in the model, suggesting that pipe wall reactions may not be important for the New Haven system. Explicit representation of reactant concentrations, temperature and pH in the kinetic model make this approach especially well suited for "what if" modeling, modeling of reactive byproducts, and modeling systems with multiple sources of different NOM quality or different chlorine doses.