Rock mass classification systems are simple but valuable tools for the qualitative and quantitative classification of rock masses and for the planning of the fortification of mining excavations. Unfor-tunately, in the prefeasibility phase, not all the information needed for a preliminary project evalua-tion is always available, and one of the few available information is the RQD, however, although it is very necessary to determine the GSI to analyze the failure criteria, it is difficult to obtain at this stage of the project. Although several correlations between the different classification systems have been identified, the most abundant ones are those relating GSI as a function of RMR and as a function of Barton's Q. As for GSI relationships as a function of RQD, only three recent relationships are available: Hoek, et al. (2013); Santa, et al. (2019); and Xia, et al. (2022). Therefore, this study presents a correlational analysis of the GSI and RQD classification systems, using robust nonparametric statistics, with the aim of determining an expression to estimate GSI in the field. Among the results, it is highlighted that better GSI prediction results are obtained when 25%<RQD≤87%, with a maximum error of ±14 points, improving the estimation accuracy by 62% with respect to current proposals. Despite the above, the difficulty of interpreting more accurately the specific geological characteristics of each rock mass remains.