In this thesis paper, the influences and performances of the curve fitting algorithms of the CT-TDLAS have been evaluated by the use of phantom data. 11 x 11 absorption lines have been installed using the phantom data, which implies 121 grids have been installed on a virtual measurement area. The phantom data distribution have been generated based upon the Gaussian distribution for temperature and concentration distribution on the virtual measurement area. The light spectra of H2O have been used for the evaluations. The phantom temperature data have been set to two temperature ranges, 300K~700K and 300K~1000K. The phantom concentration data have been set to two temperature ranges, 0.02~0.14 and 0.02~0.18. The performances of four curve fitting algorithms, One-Ratio-two-Peaks (1R2P) curve fitting, two-Ratios-three-Peaks (2R3P) curve fitting, Cross-Correlation based curve fitting, six-peaks curve fitting, have been quantitatively compared each other using the phantom data. At lower temperature range, 2R3P algorithm has shown the lowest calculation errors between the phantom data and reconstructed ones. At higher temperature range, six-peaks curve fitting algorithm has shown he lowest calculation errors. From this, it can be said that the most optimized algorithm should be adopted for the temperature ranges.