A genetic algorithm (GA) program has been developed in MATLAB to simultaneously analyse five random variables relevant to slope stability: three structural parameters (joint location, fault location and dip angle) and two strength parameters (cohesion, c and friction angle, o). The algorithm is implemented within a limit equilibrium framework and applied to a case study of bench-scale slope failure. The results indicate that the failure is structurally controlled, with the GA effectively identifying a near-maximum unstable block that corresponds closely with field observations of the failed mass. The critical block geometry derived by the GA is characterised by Joint 1 located at the toe with a lower-end dip angle and a fault positioned furthest from the bench crest, coupled with minimal cohesion and friction angle values. The analysis concludes that the failure initiated via sliding along Joint 1, where it is intersected at the rear by the fault and progressed through toppling against Joint 2, resulting in a combined sliding-toppling failure mechanism.