Shortwave downward fluxes for selected stratus, cirrus, and mixed phase cloud cases are analyzed based on cloud and surface radiation measurements from the Cloud Lidar and Radar Exploratory Test conducted in the Denver--Boulder area of Colorado during September--October, 1989. A medium resolution, discrete-ordinate shortwave radiative transfer model is used to provide clear-sky conditions and to examine the cloud shortwave radiative forcing. The model simulation indicates that for stratus clouds the effective radius increases with increasing liquid water path. For cirrus cloud simulation, the model results are within 10% agreement with the surface flux measurements. However, using the one-dimensional plane-parallel model, the model results are in poor agreement for the inhomogeneous mixed phase cloud case. Over the elevated observation site, the reduction in shortwave downward flux by clouds can be as large as 40% for a small cloud water path value of 20 g m-2. The variation in observed cloud shortwave forcing is highly correlated with the integrated cloud water path. The normalized (by the clear-sky value) cloud shortwave forcing increases rapidly when the cloud water path is small. The rate of increase decreases, and the normalized cloud forcing approaches saturation when cloud water path becomes large. The magnitude of the saturation value depends on cloud optical properties. The variation in observed cloud forcing is consistent with the theoretical curve for cloudy atmospheric albedo variation. At a constant value of cloud water path, the normalized cloud forcing increases with solar zenith angle. The solar zenith angle effect is less significant for larger value of cloud water path. ¿American Geophysical Union 1994