To meet the development needs of thorium-based nuclear energy systems for nuclear data evaluation, the aim of this study was to investigate the availability and accuracy of cross-section data for U-233 to perform calculations of the critical system for the thorium - uranium fuel cycle. By utilising the method of weighted least squares and using nonlinear regression with high-order polynomials to meet the growing demands of nuclear data, a computer programme FITWR was developed and applied to the experimental total cross-sections for MEV incident energy particles, e.g., neutrons and protons. The FITWR programme can handle variance and covariance data provided with experimental data and yield those statistical measures for evaluated data. Two evaluated data libraries are available, the U.S. data bank ENDF and the Japanese data bank JENDL. Pre-calculations were performed using the SAMMY code in conjunction with two fitting methods: the Bayes method, which is available through the SAMMY code, and the weighted least squares method with nonlinear regression.