This paper presents the use of tree-based ensemble models for predicting the shear strength of soil (SSS). The study uses a dataset of 249 soil samples and six controlling variables (depth, clay proportion, loam proportion, sand proportion, plastic limit, and plastic index) to develop an algorithm for predicting SSS. Five models were developed: extra tree regressor (ETR), decision tree regressor (DTR), ridge regression (RR), linear regression (LR), and Bayesian ridge regressor (BRR). The models were evaluated using six performance indices. The results showed that ETR outperformed the other models in terms of modeling outcomes. A sensitivity analysis was conducted to determine the importance of the influencing variables. The best model's learning process was analyzed using various plots, including learning curves, outlier distance detection, feature importance, prediction error, and residual plots. The feature significance plot revealed that sample depth is the strongest feature contributing to SSS prediction in the best model. The study highlights the effectiveness of tree-based ensemble models in predicting SSS without the need for costly laboratory testing, which is a practical requirement in geotechnical engineering. Keywords: Shear strength, Tree-based model, CV, ETR, ML.This paper presents the use of tree-based ensemble models for predicting the shear strength of soil (SSS). The study uses a dataset of 249 soil samples and six controlling variables (depth, clay proportion, loam proportion, sand proportion, plastic limit, and plastic index) to develop an algorithm for predicting SSS. Five models were developed: extra tree regressor (ETR), decision tree regressor (DTR), ridge regression (RR), linear regression (LR), and Bayesian ridge regressor (BRR). The models were evaluated using six performance indices. The results showed that ETR outperformed the other models in terms of modeling outcomes. A sensitivity analysis was conducted to determine the importance of the influencing variables. The best model's learning process was analyzed using various plots, including learning curves, outlier distance detection, feature importance, prediction error, and residual plots. The feature significance plot revealed that sample depth is the strongest feature contributing to SSS prediction in the best model. The study highlights the effectiveness of tree-based ensemble models in predicting SSS without the need for costly laboratory testing, which is a practical requirement in geotechnical engineering. Keywords: Shear strength, Tree-based model, CV, ETR, ML.