A psychophysical experiment was carried out to evaluate 5 chroma clipping techniques for gamut mapping. Three different destination gamuts were used in order to determine how the shape of the destination gamut would influence performance of the different mapping techniques. The test images consisted of 7 colored computer-rendered textured spheres and one image which contained all 7 objects. Therefore the effect of hue on the preferred technique could be gauged. Although straight chroma-clipping, keeping lightness constant, was overall the best technique, there were notable exceptions. The results indicate that trade-offs between global trends and specific exceptions to these trends may need to be implemented based on local gamut shape and object hue in the image.