Shear wave velocity in well logging is measured by DSI (Dipole Shear Sonic imager) tool. Due to high measurement costs it runs in a limited number of wells in a hydrocarbon field. From these wells, only some intervals may have Vs data. Unlike Shear wave velocity data, Compressional wave velocity data and other conventional log data such as Neutron porosity, Density and resistivity data are obtained and measured easily by related tools in well logging. These logs have mathematical and physical relations with Shear wave velocity data, and as in reservoir intervals, Shear wave velocity data are important in geophysical studies such as AVO and VSP, lithology and fluid type identification, interpretation of elastic parameters and rock mechanical properties calculation, by using conventional log data and finding logical relations between these data and Shear wave velocity data, we can estimate Shear wave velocity in equivalent and partially similar intervals or wells with no Shear wave velocity data. So, because of the importance and usefulness of shear wave velocity, in wells with no shear wave velocity log, we should use a method to predict shear wave velocity. Genetic algorithms technique as a subset of evolutionary computing is an important part of intelligent systems for solving optimization problems. In this study shear wave velocity was modeled by genetic algorithms technique from petrophysical data in Ghar member of Asmari Formation, Hendijan field. For measuring there liability of the method, predicted values were compared with the real shear wave velocity data in Ghar member of Asmari Formation, Abozar field. Comparing the estimated results from Ghar Formation in Hendijan field with real data in Ghar formation in Abozar field, we can conclude that changes in, lithology,. fluid type and other petrophysical properties can have a strong effect on Shear wave velocity variation and reliability of predicted models, and this factor shows the importance of modeling and estimation methods like GA, in verifying the effects of changes in reservoir properties in large distances and their effect on reliability of intelligent systems Multiple Regression Analysis was used as another technique to evaluate the accuracy of the optimized model with GA. The results of this study show that GA could be considered as an efficient, fast and cost effective method for predicting VS- and reducing the exploration risks in reservoir characterization studies.