While disinfection of drinking water reduces the risks of pathogenic infection, threats to human health due to the formation of disinfection byproducts (DBPs) may arise due to natural organic precursors. Regression-based models characterizing the formation of DBPs are derived from data for 28 conventional water treatment plants in Ontario. DBPs are shown to be correlated statistically with dissolved organic carbon, pre-and post-chlorination dosages, pH and temperature. Using backward elimination nonlinear regression, a set of mathematical functions are obtained (R2=0.62 to 0.79) for an array of DBPs. The models are used to guide decision-markers in the selection and operation of drinking water treatment processes to decrease DBP formation, indicating that a shift from emphasis on pre-chlorination to post-chlorination has the most effect on DBP formation.