Vicarious calibration and field validation is a critical aspect of NASA's Earth Observing System program. As part of calibration and validation research related to this project, the Remote Sensing Group (RSG) of the Optical Science Center at the University of Arizona has developed an imaging radiometer for ground-based measurements of directional reflectance. The system relies on a commercially available 1024¿1024 pixel, silicon CCD array. Angular measurements are accomplished using a fish-eye lens that has a full 180¿ field of view with each pixel on the CCD array having a nominal 0.2¿ field of view. Spectral selection is through four interference filters centered at 470, 575, 660, and 835 nm. The system is designed such that the entire 180¿ field is collected at one time with a complete multispectral data set collected in under 2 min. The results of laboratory experiments have been used to determine the gain and offset of each detector element as well as the effects of the lens on the system response. Measurements of a stable source using multiple integration times and at multiple distances for a set integration time indicate the system is linear to better than 0.5% over the upper 88% of the dynamic range of the system. The point spread function (PSF) of the lens system was measured for several field angles, and the signal level was found to fall to less than 1% of the peak signal within 1.5¿ for the on-axis case. The effect of this PSF on the retrieval of modeled BRDFs is shown to be less than 0.2% out to view angles of 70¿. The degree of polarization of the system is shown to be negligible for on-axis imaging but to have up to a 20% effect at a field angle of 70¿. The effect of the system polarization on the retrieval of modeled BRDFs is shown to be up to 3% for field angles of 70¿ off nadir and with a solar zenith angle of 70¿. Field measurements are made by mounting the camera to a boom mounted to a large tripod that is aligned toward south. This tripod obstructs sampling of the surface reflectance past 25¿ off nadir northward. The system is typically operated at a height of 1.5 m to view over a large sampling of surface features, such as cracks. To evaluate the surface BRDF, measurements are collected throughout the morning as a function of Sun angle. A single measurement consists of all four bands and a dark-current measurement. Data sets have been collected over several vicarious calibration sites and calibration tarpaulins. Comparisons with measurements made by a simple goniometer-based system indicate that the camera system is as accurate as the goniometer. Scattering phase function values derived from the camera system are fit to a modified Pinty-Verstraete equation. This function is shown to fit the data to better than 0.3% for data collected during an example RSG vicarious calibration experiment. Bidirectional reflectance data derived from the camera system also compare well to those predicted from the Walthall model. These BRDF models are critical for determining the applicability of measurements taken over small areas to represent the BRDF properties of an entire site, which in some cases is of the order of several kilometers in size. ¿ 2001 American Geophysical Union