Abstract: Objectives: Acoustic ray-tracing method is an important means to solve the problem of acoustic ray bending in the process of propagation. It was used to calculate the slant range of deep-water object positioning can effectively attenuating the influence of acoustic ranging system error. However, the accompanying problem is the reduction of computational efficiency. To solve this problem, we put forward a self-adaptive layering method based on the area difference of sound velocity profile (SVP), which reduces the computation by optimizing the layering strategy of the SVP. Methods: Firstly, the relationship between the SVP area difference and ranging error is established according to the research of thirteenth reference. Based on this relationship, the constant-gradient and the zero-gradient ray-tracing method were analyzed which is more suitable for slant range calculation. And the maximum tolerance of the SVP area difference is obtained by setting the maximum tolerance of ranging error. Secondly, the structural layering is carried out according to the change law of sound velocity gradient, and the refined layering is carried out on constraint of the maximum tolerance of SVP area difference. Results: The results show that: (1) The measured SVP in the same sea area and during a similar time can be considered as the same cluster SVP, which satisfies the Eq. (1). The average value of multiple measured SVPs can be approximately considered as the background SVP to estimate the linear coefficient k0. (2) The adaptive layering method can optimize the layering scheme according to the changing law of the SVP curve. The layering interval will increase where the gradient change rate is small and reduce the layering interval where the gradient change rate is large, so as to reduce the number of layers as much as possible under the condition of meeting the maximum tolerance ranging error. (3) Compared with equally spaced layering 10m, when the number of layering is reduced by 86%, the ranging error caused by layering is lowered from centimeter level to millimeter level, which proves the effectiveness of this method. Conclusions: The adaptive method has strong robustness and practicability. It can adjust the layering strategy according to the structural characteristics of SVP and usage scenarios of the offshore operation. Since the number of layers of the SVP is greatly reduced, the calculation speed is greatly improved. This will be helpful for large amount of data process or real-time underwater acoustic navigation.