<p>Accurately identifying the most probable point (MPP) is essential in structural reliability analysis, as the region around the MPP usually contributes the most to the failure probability. In high-dimensional spaces, the curse of dimensionality causes the search domain to grow exponentially, which increases the complexity of MPP identification and highlights the need for robust MPP estimation methods. In this study, we propose a sensitivity analysis-enhanced cooperative competitive krill herd (SA-CCKH) algorithm for high-dimensional MPP search. SA-CCKH couples the global search of the cooperative competitive krill herd method with a sensitivity analysis mechanism that applies locally weighted perturbations to accelerate convergence and reduce computational cost. Once identified, the MPP serves as the basis for subsequent analysis. In this study, it is employed in conjunction with importance sampling to perform reliability analysis. The efficiency and accuracy of SA-CCKH are demonstrated on three numerical benchmarks and a practical engineering example.</p>