The aim of this article is to present the evaluation of gamut mapping algorithms (GMAs) in a series of three experiments intended to serve as the basis fur developing solutions that are accurate and universally applicable. An evolutionary gamut mapping development strategy is used, in which Jive test images are reproduces between a CRT and printed media obtained using different GMAs. Initially: a number of previously published algorithms were chosen and psychophysically evaluated, whereby an important characteristic of this evaluation was the separate evaluation for individual colour regions within test images. New, algorithms were then developed on this experimental basis, subsequently evaluated, and the process was repeated once more. In this series of experiments, the new GCUSP algorithm, which consists of a chroma-dependent lightness compression followed by a compression? towards thee lightness of the reproduction cusp on the lightness axis, gave the most accurate and stable performance overall. The results of these experiments were also useful for improving the understanding of some gamut mapping factors-in particular gamut difference between media. (C) 2001 John Wiley & Sons, Inc.