This work introduces a new approach for probabilistic elastoplastic topology optimization based on the improved bidirectional structural optimization (BESO) technique. To consider uncertainties, the volume fraction and the material properties are considered randomly. Thus, the reliability-based design is integrated into the deterministic design by applying a reliability constraint to the optimization problem. Furthermore, using limit analysis, a bound is applied to the plastic limit load multipliers to govern the plastic behavior of the problems. Results from a 2D benchmark problem are used to illustrate how adequate the approach that has been provided is. Also, a 2D elastoplastic numerical example is shown to illustrate the proposed method’s capability of identifying the best topology for elastoplastic models in the context of reliability-based design. The results indicate that the reliability constraints work effectively as a bound that reduces the yielding states.