The application of genetic algorithms (GA) optimization approach supported by the first-order derivative (FOD) and Newton-Raphson (NR) methods to time-resolved polarized fluorescence spectroscopy, is discussed. It is demonstrated that the application of both methods to x(2) function reduces the number of adjustable model parameters. The combination of GA-optimizer with the FOD and NR methods improves considerably the efficiency of global analysis of kinetic and polarized fluorescence decays for solutions and organized media, including the case of excited-state processes. © 2005 Elsevier B.V. All rights reserved.