The formation of organochlorine compounds in drinking water in the water supply system chlorination process was studied. The dynamic pattern of chloroform concentration in drinking water in a water treatment facility over a year is provided. Cancer health risk for the population of using drinking water is calculated. A greater speed of organochlorine compound formation in comparison with the rising concentration of residual chlorine associated with a bigger dose of chlorine agent was documented. The influence of such factors as pH medium, permanganate oxidizability, chlorine absorbing capacity, temperature and residual chlorine concentration on the concentration of chloroform in drinking water was experimentally established. A neural network model that predicts the chloroform concentration in drinking chlorine treated water was used. The chlorine agent dosage control organization chart was proposed on the basis of fuzzy logic that can be used in design of water-supply systems taking into account the health risks for the population.