Biological tissues undergo complex phase change heat transfer processes during cryosurgery, and a theoretical model is preferable to forecast this heat experience. A mathematical model for phase change heat transfer in cryosurgery was established. In this model, a fractal treelike branched network was used to describe the complicated geometrical frame of blood vessels. The temperature distribution and ice crystal growth process in biological tissue including normal tissue and tumor embedded with two cryoprobes were numerically simulated. The effects of cooling rate, initial temperature, and distance of two cryoprobes on freezing process of tissue were also studied. The results show that the ice crystal grows more rapidly in the initial freezing stage (<600 s) and then slows down in the following process, and the precooling of cryoprobes has no obvious effect on freezing rate of tissue. It also can be seen that the distance of 10 mm between two cryoprobes produces an optimal freezing effect for the tumor size (20 mm × 10 mm) in the present study compared with the distances of 6 mm and 14 mm. The numerical results are significant in providing technical reference for application of cryosurgery in clinical medicine.