This study examined chlorite and chlorate formation resulting from statistically-designed experiments performed using both "synthetic" and natural waters which were dosed with chlorine dioxide. The purpose was to screen various physical/chemical parameters such that those having significant effects on chlorite and chlorate formation would be further investigated during model development experiments using raw waters. Also, it allowed the assessment of whether 2-factor parameter interactions would be considered when developing the predictive disinfection byproducts model equations. Predictive equations were assessed using both internal and external validations. Sensitivity analyses consisted of examining parameter contributions to the "explanatory power" of the relevant model equations. Model equations were developed which were capable of predicting chlorine dioxide consumption, and chlorite and chlorate formation within the boundary conditions which were examined, while maintaining reasonable justification and a statistically-sound base.