This paper develops an effective and new method to solve a class of fractional differential equations. The method is based on a fractional order reproducing kernel space. First, depending on some theories, a fractional order reproducing kernel space W [0, 1] is constructed. The fractional order reproducing kernel space is a very suitable space to solve a class of fractional differential equations. Then, we calculate the reproducing kernel R y (x) of the space W [0, 1] skilfully in §3. And convergence order and time complexity of this algorithm are discussed. We prove that the approximate solution u n of (1.1) converges to its exact solution u is not less than the second order. The time complexity of the algorithm is equal to the polynomial time of the third degree. Finally, three experiments support the algorithm strongly from the aspect of theory and technique. KEYWORDS complexity analysis, convergence order, fractional reproducing kernel space, initial value problem of fractional equations MSC CLASSIFICATION 34A45; 65L05; 65L20 1 INTRODUCTION Fractional differential equations(FDEs) have important applications in signal processing, physics, controller design, electric circuits, and so forth. Fractional order modelling minimizes the inaccuracy that arises from the ignorance of significant real parameters. It permits a greater degree of freedom in the model compared to an integer-order system. Thus, in the last two decades, the theory and numerical analysis of FDEs are receiving more and more attention. Many schemes such as fractional linear multistep method, 1 operational matrix method, 2,3 homotype analysis method, 4,5 moving least square reproducing kernel method, 6 predictor-corrector method, 7 dual reciprocity boundary integral equation technique, 8 finite difference method, 9 Adomian decomposition method, 10 extrapolation method, 11 Chebyshev wavelets method, 12 modified fractional Euler method, 13 variational iteration method, 14 and Laplace transform 15 have been developed to determine the numerical solutions of several classes of fractional ordinary differential equations and partial differential equations (PDEs). Recently, Kumar et al 16 used Bernstein wavelet and Euler methods to handle nonlinear fractional predator-prey biological model. In 2020, Kumar et al 17 utilized residual power series and perturbation methods to handle Fokker-Plank equation. Veeresha et al 18 used the q-HATM to acquire the solution of FGNLS equation. Kumar et al 19 applied new amagalation of homotopy analysis method, Laplace transform method, and Adomian polynomial's for fractional discreate KdV equations. In addition, Kumar et al 20 employed homotopy perturbation transform method to obtain approximate analytical solutions of the time-fractional PDEs. Odibat et al 21 analyzed systems of nonlinear FDEs by using homotopy asymptotic method (HAM). Kumar et al 22 suggested two new fractional derivatives, namely, Yang-Gao-Tenreiro