The surface profiles of retrograded rice starch (RS) samples stored at different times were obtained by horizontal sectioning and scanning electron microscopy (SEM). SEM images showed that the surface topography of the retrograded RS samples possessed fractal characteristics, which was also proven by fractal analysis. The fractal features for analysis were extracted using a new image processing method. The average fractal dimensions of the retrograded RS samples stored for 1, 5, 10, and 15 days were 1.6587, 1.7333, 1.7807, and 1.8340, respectively. The method indicated that the fractal dimension increased as the extent of retrogradation increased. The good correlation between fractal dimensions and the retrogradation enthalpies of the retrograded samples were established using a fitted binomial model (R(2) = 0.9976). These results show that the fractal dimensions obtained using this new image processing method could effectively quantify the extent of retrogradation. (C) 2011 Elsevier Ltd. All rights reserved.