This article proposes Gamma-Compression Gamut Mapping Algorithm (GMA) based on the concept of Image-to-Device. Considering the relations of color gamut boundaries between the devices (such as printer, monitor) and image source, the proposed method adjusts the image color distributions, fitting to the output device gamut. The proposed image-to-device GMA works to preserve its gradation and chroma with minimum losses depending on the color distributions in the segmented hue-leaves. The following three typical mapping methods are compared and discussed: (1) Device-to-Device Gamma-Compression GMA, (2) Image-to-Device Gamma-Compression GMA; (3) Clipping GMA. Two kinds of color spaces, CIEL*a*b* and CIECAM97s-JCh, are used for testing these GMAs. The psychophysical experiments in CIECAM97s-JCh space resulted in better rendition than did CIEL*a*b* space. In natural color imaging applications, the proposed Image-to-Device GMA is superior to conventional Deviceto-Device GMA, and is a better way to maintain its gradation and higher chroma.