Conventional water treatment plants in Nigeria, such as the Ede Water Works, often rely on static jar testing for chemical dosing, a method poorly suited for handling dynamic raw water quality. These approach had resulted in sub-optimal coagulation and disinfection, increasing costs and risks to public health. This study optimized coagulant (alum) and disinfectant (chlorine) dosages for conventional surface water treatment for Ede Water Works. Water samples were collected from raw, aerated, and treated stages during dry and rainy seasons. Key parameters (pH, turbidity, chemical Oxygen demand (COD), biochemical oxygen demand (BOD) alkalinity, microbial counts) were analysed. Jar tests were conducted using alum concentrations of 50-100ppm, and chlorine dosing was performed for microbiological removal. The model identified optimal alum and chlorine dosage of 60-70ppm and 70-80ppm could effectively reduce turbidity and achieved complete microbial inactivation. The RSM models demonstrated strong predictive power, with R² values of 0.827, 0.957, and 0.958 for turbidity, COD, and BOD, respectively. The study concluded that replacing reactive jar testing with a proactive, model-based dosing strategy significantly enhances treatment efficiency, ensures regulatory compliance, and reduces operational costs, offering a viable framework for improving water treatment in similar contexts.