In the present research, the photocatalytic degradation of dimethyl sulphoxide (DMSO) by means of CdS/TiO2 core/shell nanocomposite was investigated and modelled for the direct measurement of degradation rate. Instead of common measurement methods, liquid freezing point measurement was used in order to determine the rate of DMSO decomposition. To model the photocatalytic behaviour, two empirical-statistical equations based on photocatalytic retention time, amount of catalyst and pollutant were introduced. The effect of parameters such as retention time, amount of catalyst and pollutant were studied by statistical methods. The design of experiments, acquisition and optimization of statistical models was performed by response surface methodology (RSM) through central composite design (CCD). The rates of disappearance fitted the Langmuir-Hinshelwood kinetics model and the parameter k was determined to be up to 0.0105 per minute for a low concentration and 0.0073 per minute for a high concentration of DMSO. In addition, more than 85% degradation of 1% DMSO was attained by 8% catalyst in 150 minutes. Finally, the accuracy and consistency of the statistical models was verified by the HPLC method and +/- 2.7% difference was observed. The other results indicate that the CdS/TiO2 nano photocatalyst can efficiently remove dimethyl sulphoxide from wastewater under visible light irradiation.