Stellar populations of high redshift radio galaxies (HzRG) (z up to 4.2) are the oldest stellar systems known, that is the ones formed at the earliest cosmological epochs. Therefore they are the best objects for providing us with information about the epoch of galaxy formation. The information on the stellar populations in HzRG are obtained from the study of their Integrated Spectral Energy Distribution (ISED) which are gathered both from spectra and integrated magnitudes. The most common approach for the interpretation of colors and spectral features of the energy distribution of galaxies is the Evolutionary Population Synthesis (EPS), which has been introduced for the first time by Tinsley in 1972. EPS models have often been used in the past to interpret the ISED of HzRG (Chambers & Charlot 1990; Lilly & Longair 1984; di Serego Alighieri et al. 1994) in order to draw conclusions on the age of the stellar populations and therefore on the epoch of galaxy formation. The results are sometimes conflicting and a number of very recent EPS models have become available (Bressan et al. 1995; Bruzual & Charlot 1993; Buzzoni 1989; Guiderdoni & Rocca-Volmerange 1987): we are therefore analysing the differences between the various EPS models with the aim of assessing their suitability to study the stellar population at early epochs. The EPS models assume for stars a given Initial Mass Function(IMF) as well as a Star Formation Rate (SFR). Then one can compute the number of stars with given mass present in the galaxy as a function of time. The position of each star in the HR diagram is determined by means of the isochrones, which are calculated from stellar evolutionary models. The ISED of a galaxy is obtained from the superposition of the spectra of single stars obtained from a stellar spectral library. Thus these models describe the galaxy ISED as a function of the time, giving a complete evolutionary picture.