As a ground improvement technique, stone columns are often used in soft soil to increase the bearing capacity and reduce total and differential settlements. These columns act as vertical drains to speed up consolidation and decrease settlement. In the design and construction of stone columns, uncertainties may arise from variations in soil properties, changes in loading conditions, and the execution of the installation process. Keeping in view the uncertainties associated with soil and stone column parameters, reliability analysis of the bearing capacity of stone columns has then been carried out using three methods of reliability analysis, namely Mean-Value First-Order Second Moment (MVFOSM) method, First-Order Reliability Method (FORM) and Monte Carlo Simulation Method (MCSM). The basic computational codes for estimation of the bearing capacity of stone column improved soil based on the Gibson and Anderson cavity expansion theory and the relevant IS code [13] have been developed in the MATLAB environment. The effect of the probability distribution assumption, the impact of the coefficient of variation (COV) of random variables, and the effect of correlation among the random variables are also studied. Upon reviewing the results, it has been noted that despite the deterministic analysis indicating a factor of safety (FS) for the bearing capacity of over 1.5, the reliability studies reveal a 20% probability of failure. This observation supports the necessity of employing a probabilistic approach for analysis.