The emerging need for a fast, safe economical approach to global and localized measures of desaturation of hemoglobin with oxygen (HbO2) in the human brain motivates further research on time-resolved spectroscopy in four areas of study. (1) To afford quantization of hemoglobin saturation through time-resolved spectroscopy in the time domain (TD) and in the frequency domain (FD). Evaluation of dual-wavelength TD and FD spectrometers for determining quantitatively hemoglobin desaturation and blood-volume changes by calculations that are insensitive to mutual interference is proposed. The diffusion equation, as it applies especially to TD studies, and the absorption ((mu) a) and scattering ((mu) s) coefficients provide their independent determination from the late and early respective portions of the kinetics of the emergent photons in response to a short input pulse (50-100 psec). (2) The identification of the photon-pathlength change due to the arterial pulse in the brain tissue by FD methods with Fourier transformation affords an opportunity to employ principles of pulse oximetry to vessels localized deep within the brain tissue. (3) Localization of desaturation of hemoglobin in portions of the brain can be achieved through dual-wavelength scanning of the input/output optical fibers across the head for an X-Y coordinate and varying the distance between input and output ((rho) ) or the time delay in data acquisition to afford an in-depth Z scan. Localizations of shed blood, which have an effective concentration of over 10 times that of capillary-bed blood, are identified by X, Y, Z scans using only a single wavelength. (4) Independent measurements of absorption ((mu) a) and scattering ((mu) s) coefficients, particularly by TD techniques, affords structural mapping of the brain, which can be used to diagnose brain tumor and neuronal degeneration. Two experimental systems are used to critically evaluate these studies; the first, a hemoglobin/lipid/yeast model in which intermittent oxygenation gives saturation/desaturation effects and addition of hemoglobin simulates increased blood volume. These models can be global or may contain localized ''black'' absorbers simulating brain bleeds or model-stroke volumes in which oxygenation/deoxygenation simulates normoxia/hypoxia. Secondly, animal brains are used to model the following changes in vivo: global or localized hypoxia, brain bleeding, and hematomas by epidural blood injection, and physiological changes by epilepsy. Neuronal degeneration causing scattering effects is modeled by injection, epidurally or into the animal model brain, highly scattering material such as polystyrene spheres. The proposal envisages a basic science study of photon migration in the brain with important applications to stroke, epilepsy, brain trauma, and neuronal degenerative disease.