Abstract: To slove the problem of insufficient driving torque during the slope-crossing of pure electric vehicles, a control strategy was proposed based on road slope with acceleration pedal change rate as input and compensation torque as output. A slope recognition algorithm was adopted to effectively identify the real road slope. Vehicle model and control algorithm model were established in Matlab/Simulink for simulation analysis. The vehicle acceleration and climbing performance was simulated under 4% slope condition, and the speeds under the two control strategies were compared. The results show that the torque optimization control strategy can better identify the driver′s driving intent. For the optimized torque control strategy, the acceleration time is 10.65 s at the speed from 0 to 50 km·h^-1, which is 11.62% lower than that of the basic torque control strategy. The uphill acceleration time is 8.60 s at the speed from 50 to 80 km·h^-1, which is 14.85% lower than that of the reference torque control strategy. The optimal torque control strategy can effectively improve the acceleration and climbing performance.