The advancement in spectral analysis methods for the emission spectrum of ruby has been driven by the characterization of R-line peak shifts with stress in order to establish piezospectroscopic relationships. These relationships form the basis for the development of photo-stimulated luminescence spectroscopy (PSLS) as a nondestructive method to determine the integrity of the thermally grown oxide (TGO) layer on jet engine turbine blades. Besides the measurement technique, the accuracy of PSLS in stress measurements is influenced by the spectral analysis methodology, which is the focus of this paper. Gradient-based algorithms have been used widely in the methods developed thus far. The approach of using genetic algorithms in the spectral analysis of R-lines and vibronic bands is presented here for the first time and validated with the wellknown piezospectroscopic coefficients of the R-lines. The implementation of this method has led to significant new results in the quantification of peak shifts with uniaxial stress in the vibronic bands of the spectrum. The use of genetic algorithms is instrumental in the deconvolution and fitting of the numerous peaks in these bands. Fitting statistics, such as the fitness function and number of function evaluations, were used to assess the effectiveness of the procedures used in this method.