Abstract: Automatic test data generation is a key link in the process of test automatic technology.In order to improve the efficiency of testing automation, a new algorithm was proposed to improve the traditional cuckoo search.Chaotic sequence generated by cube map was used to initiate individual position, which strengthened the diversity of global searching.And a Gaussian disturbance would be given on the local optimum of each generation, which could increase the vitality of population change and accelerate the convergence rate.At the same time, the algorithm is also designed to fit the fitness function to accelerate the optimization process of the data.Through the triangle, mid and equal program, the basic cuckoo search and the basic genetic algorithm were compared.Experiments result shows that the new algorithm has obvious advantages in efficiency and effectiveness compared with the other two for test case generation.