In order to improve the performance of image enhancement, effectively preserve original image details, an image enhancement method using multi-scale gray-level transformation was presented. The original image was decomposed into multi-scale layers using gradient domain based recursive bilateral filtering. Based on wavelet transform, the subcinctures of decomposed layers were transformed respectively in the gray-level. The overall image enhancement was realized based on the enhancement results of decomposed layers, which were reconstituted with the transformed subcinctures. Compared with histogram equalization and gray-level transformation, the proposed method performed better in image enhancement, and preserved details. The performance of the enhancement results was evaluated by the objective indicator. The experimental results show that the proposed method has excellent effect, simple structure and lower computational complexity.