Ladar, which can obtain range image and intensity image, may provide much useful information than other imaging system. Range image noise suppression is a key for ladar image processing. The suppression effects influence directly precision in tracking and pointing. According to range image noise, combining intensity image with multilevel median filtering is used for actual ladar image. First, dropout is removed according to intensity image mean in 3×3 window and whole image mean. Then, outlier is removed by using multilevel median filter. Each processing result is presented and better range image is obtained. The investigation results show that this algorithm can suppress range image noise in coherent ladar effectively.