A genetic algorithm (GA)-based method has been developed to analyze Charge-eXchange Recombination Spectroscopy (CXRS) data and provide in-between shot evaluation of the ion temperature profile during Tore Supra experiments. The GA method proposed here proves to be fast and fairly accurate, even when analyzing low signal-to-noise data. Simulations using theoretical signals suggest that the ion temperature and the plasma rotation velocity are expected to be determined with a precision better than 10% for a noise level up to 5% of the spectrum peak. The good agreement with the commonly used KS4FIT code when analyzing CXRS experimental data - typically within 30% for ion temperature measurements - confirms the efficiency of such an analysis tool.
