This paper presents a genetic algorithm (GA) based method for bicriterion operating optimisation in DC railway power systems, in order to achieve a tradeoff between high power recovery and uniform load sharing among traction substations (TSS). Based on a multi-train movement simulation and line receptivity study, a scheme for optimally coordinating control of firing angles on the TSS is proposed. The developed method can be used to any operating conditions, but the emphasis of this paper is on regenerative braking in DC railways. Simulation results have shown that much improvement on power recovery and load sharing can be achieved by using the proposed method