Usually, an UAV(Unmanned Aerial Vehicle) path planning problem is abstracted into a nonlinear optimal control model, that is, a nonlinear programming. However, this kind of problem is always difficult to obtain stable solutions quickly. In this paper, a sequential convex programming whose cost function and inequality constraints are convex is proposed to approximate the nonlinear programming to obtain the solution. Under mild condition, the solution sequence generated by the sequential convex programming is convergent to the KKT(Karush-Kuhn-Tucker) point of the origin problem, which has been verified by a rigorous theoretical proof in this paper. Thus, a nonlinear programming can be solved by a series of sequential convex programming. The effectiveness of the proposed method is verified by an UAV path planning application.