A method for processing a radiographic image, containing: a step for calculating an average S.sub.m of original image data of an averaging area around each pixel of the radiographic image; and a step for transforming the original image data S of each pixel of the radiographic image to obtain a processed image data Q of each pixel in accordance with an equation, Q=S+.function.(S-S.sub.m), where .function.(S-S.sub.m) has a characteristic in at least one of first and second ranges of the difference S-S.sub.m, the difference S-S.sub.m is positive in the first range and negative in the second range, and according to the above characteristic, the absolute value .vertline.Q-S.vertline. varies in accordance with an increasing monotonic function of the absolute value .vertline.S-S.sub.m .vertline. when the absolute value .vertline.S-S.sub.m .vertline. is less than a predetermined value, and the absolute value .vertline.Q-S.vertline. varies in accordance with a decreasing monotonic function of the absolute value .vertline.S-S.sub.m .vertline. when the absolute value .vertline.S-S.sub.m .vertline. of the difference S-S.sub.m is more than the predetermined value. The above function .function.(S-S.sub.m) may be a non-odd function of the difference S-S.sub.m.