The term has various meanings it goes by. In color reproduction and colorimetry, a gamut, or color gamut /ˈɡæmət/, is a convex set containing the colors that can be accurately represented, i.e. reproduced by an output device (e.g. printer or display) or measured by an input device (e.g. camera or visual system). Devices with a larger gamut can represent more colors. Similarly, gamut may also refer to the colors within a defined color space, which is not linked to a specific device. A trichromatic gamut is often visualized as a color triangle. A less common usage defines gamut as the subset of colors contained within an image, scene or video.

The term gamut was adopted from the field of music, where the medieval Latin expression "gamma ut" meant the lowest tone of the G scale and, in time, came to imply the entire range of musical notes of which musical melodies are composed. Shakespeare's use of the term in The Taming of the Shrew is sometimes attributed to the author / musician Thomas Morley. In the 1850s, the term was applied to a range of colors or hue, for example by Thomas de Quincey, who wrote "Porphyry, I have heard, runs through as large a gamut of hues as marble."

The gamut of a device or process is that portion of the color space that can be represented, or reproduced. Generally, the color gamut is specified in the hue–saturation plane, as a system can usually produce colors over a wide intensity range within its color gamut. Device gamuts must use real primaries (those that can be represented by a physical spectral power distribution) and therefore are always incomplete (smaller than the human visible gamut). No gamut defined by a finite number of primaries can represent the entire human visible gamut. Three primaries are necessary for representing an approximation of the human visible gamut. More primaries can be used to increase the size of the gamut. For example, while painting with red, yellow and blue pigments is sufficient for modeling color vision, adding further pigments (e.g. orange or green) can increase the size of the gamut, allowing the reproduction of more saturated colors.

While processing a digital image, the most convenient color model used is the RGB model. Printing the image requires transforming the image from the original RGB color model to the printer's CMYK color model. During this process, the colors from the RGB model which are out of gamut must be somehow converted to approximate values within the CMYK model. Simply trimming only the colors which are out of gamut to the closest colors in the destination space would burn the image. There are several algorithms approximating this transformation, but none of them can be truly perfect, since those colors are simply out of the target device's capabilities. This is why identifying the colors in an image that are out of gamut in the target color space as soon as possible during processing is critical for the quality of the final product. It is also important to remember that there are colors inside the CMYK gamut that are outside the most commonly used RGB color spaces, such as sRGB and Adobe RGB.

Color management is the process of ensuring consistent and accurate colors across devices with different gamuts. Color management handles the transformations between color gamuts and canonical color spaces to ensure that colors are represented equally on different devices. A device's gamut is defined by a color profile, usually the ICC profile, which relates the gamut to a standardized color space and allows for calibration of the device. Transforming from one gamut to a smaller gamut loses information as out-of-gamut colors are projected on to the smaller gamut and transforming back to the larger gamut does not regain this lost information.

Colorimetry is the measurement of color, generally in a way that mimics human color perception. Input devices such as digital cameras or scanners are made to mimic trichromatic human color perception and are based on three sensors elements with different spectral sensitivities, ideally aligned approximately with the spectral sensitivities of human photopsins. In this sense, they have a similar gamut to the human visual system. However, most of these devices violate the Luther condition and are not intended to be truly colorimetric, with the exception of tristimulus colorimeters. Higher-dimension input devices, such as multispectral imagers, hyperspectral imagers or spectrometers, capture color at a much larger gamut, dimensionally, than the human visible gamut. To be perceived by humans, the images must first be down-dimensionalized and treated with false color.

The extent of color that can be detected by the average human, approximated by the standard observer, is the visible (or visual) gamut. The chromaticities present in the visible gamut are usually visualized in the CIE 1931 chromaticity diagram, where the spectral locus (curved edge) represents the monochromatic (single-wavelength) or spectral colors. As current displays have a smaller gamut than the visible gamut, the colors that are out-of-gamut are reproduced as colors inside the display's gamut. Device gamuts are generally depicted in reference to the visible gamut. The standard observer represents a typical human, but colorblindness leads to a reduced visible gamut.

Optimal colors are the most chromatic colors that surfaces can have*. The color solid bounded by the set of all optimal colors is called the optimal color solid or Rösch–MacAdam color solid. For now, we are unable to produce objects with such colors, at least not without recurring to more complex physical phenomena.