Liquid crystal displays (LCDs) are currently the leading candidate technology for the next generation of high-performance color workstations. Color LCD technology has been maturing at a rapid pace along with major improvements in display performance. This is at least in part due to large research and development expenditures focused on bringing color LCD technology to market. LCDs offer the extraordinary design flexibility of a light valve based device, while exhibiting the desirable characteristics of relatively low volume, weight, and power consumption. Moreover, with proper optimization of optical and electronic components, color LCDs are capable of color per-formance equal to or exceeding that of the venerable shadow-mask color CRT.For a color CRT, the spectral composition and intensity of emitted light is a function of the excitation of phosphors by a stream of electrons emitted by a cathode. Since the intensity of emissions from each of the three primary color phosphors (R, G, and B) is directly proportional to the beam current while the spectral composition of the emission from each phosphor is invariant across beam current, the color and luminance of the color CRT are simply related to the luminous proportions of emitted light from each of the three primary color phosphors. There are generally no active optical elements involved, and the spectral composition of the output of the color CRT can be considered to be isotropic as well as homogenous across space and time. Thus, assuming that the spectral power distribution of the emissions from the R, G and B phosphors are known and suitably converted to tristimulus values (e.g. via the CIE 2° color matching functions), the colorimetric and photometric performance of the CRT may be readily characterized andoptimized in tristimulus space.Unlike the color CRT, which is optically simple and thus relatively easy to model and optimize in terms of colorimetric/photometric performance, the color LCD is optically complex. A transmissive color LCD is composed of a source of illumination and a multitude of layered optical elements which each modify the spectral composition of light originating from the source. Moreover, some of these elements, such as polarizers, retardation films and the liquid crystal (LC) layer itself, are optically anisotropic and birefringent layers which produce complex spectral modifications that vary as a function of the material parameters and construction of the LC cell, display voltage (i.e., luminance or gray level), and the direction of light propagation.It should be apparent that the colorimetric modeling and optimization of an LCD is a much more complex task than comparable analyses for a color CRT. In this paper we describe the foundations of LCD operation, including some basics of LCD optical models and the effects of various LCD design parameters on the spectral, intensive, and angular propagation of light through an LCD. Further, we present a method for estimating the colorimetric and photometric characteristics of color LCDs which enables the investigation of the effects of variations in a number of critical color LCD components (e.g., spectral power distribution of the source of illumination, color filter dye concentration and thickness, LC birefringence, and LCD cell construction geometry) on the ultimate chromaticity and luminance rendering capabilities of the display. Finally, we present both modeled and empirical data on the color performance of today's state-of-the-art color LCDs, compare this performance with that of existing color CRTs, and discuss the promises and problems of this new generation of high-performance color imaging devices.