We describe a new edge detector based on the robust rank-order (RRO) test, which is a useful alternative to the Wilcoxon test, for detecting edges in noisy images. Our method is based on detecting local changes in gray level intensity between adjacent pixel neighborhoods using an edge-height model to extract exact edges from images corrupted with noises. Some experiments of statistical edge detectors based on the Wilcoxon test and T test, and the well-known Canny detector and Sobel detector with our RRO detector are carried out on synthetic and real images corrupted by both Gaussian and impulsive noises. The results show that the performance of the proposed edge detector appears to be the most robust to variations in noise, performing well in all noise distributions tested.