Dai, R.-H., Zhang, Y.-P., Zhang, F.-C. and Yin, C., 2121. Self-adaptive edge-preserving smoothing and its applications in seismic impedance interpretation. Journal of Seismic Exploration, 30: 303-318. Seismic attributes, such as seismic impedance, AVO or AVA attributes, or other amplitude-like attributes, and so forth, increase the geological information interpretation ability of seismic data. However, in practical case, the calculation of seismic attributes is based on the mathematical formula, such as derivative or integration of seismic data. So, it also enhances random noise. Edge-preserving smoothing (EPS) method can suppress random noise along reflectors while preserving major stratigraphic or structural discontinuities features. These features are very important for seismic data’s geological interpretaion. However, the conventional EPS filter use fixed filter window size to perform in practice. Hence, the little geological features (e.g., channels, minor fault or thin layers) will be suppressed if their width are smaller than used filter window size. On the other hand, if the filter window size is too small, noise will not be removed sufficiently. In order to overcome this issue, we present a new EPS filter which uses a series of different window size and self-adaptively chooses the best one through filter window size scanning. The self-adpative EPS filter can strike a balance between noise remove and useful geological information protection. Applications on model tests and real data examples have shown the effectivity of the proposed method.