A method of path planning for array-based pick-and-place performed with a robotic arm is characterized in that: during each instance of the pick-and-place process performed with the robotic arm, an X-axis position sensor and a Y-axis position sensor sense coordinate errors of a pick-and-place point such that a controller calculates a position compensation value according to the sum of vectors of the coordinate errors, corrects the pick-and-place position of the robotic arm according to the position compensation value, and generates the coordinates of the next pick-and-place point. By repeating the aforesaid process flow, it is feasible to perform plenty array-based pick-and-place jobs.