The Earth Radiation Budget Experiment (ERBE) used angular direction models (ADMs) to convert satellite scanning radiometer measurements to radiative fluxes at the top of the atmosphere. The ADMs were assumed to depend only on the physical characteristics of the scene being observed, and they were taken to be independent of the instrument's spatial resolution. As is shown here, however, the angular distribution of the radiation fields derived from ERBE observations depends on the field of view size. Because the spatial resolution of the field of view depends on the satellite view zenith angle, the dependence of the angular distribution on field of view size suggests that the radiative fluxes derived from ERBE have biases which depend on satellite view zenith angle. ERBE scanner observations from the Earth Radiation Budget Satellite (ERBS) are averaged for nadir pixels to construct observations that have a constant field of view size for each of the ERBE satellite view zenith angle bins. The angular distribution of the radiation field for the constant size field of view observations exhibits systematic differences of the order of 2--5% from that obtained with the unaltered, full resolution scanner observations. While the differences are small, they are frequently significant at the 90% confidence level, given the variation in the observations and the sample size (3 months). Reflected radiances constructed from the constant size fields of view are generally more anisotropic than those constructed from the full resolution scanner observations. The population of scene type obtained for the constant size fields of view shows that the ERBE scene identification algorithm has to be modified in order to develop angular direction models which are consistent with the observed anisotropy and also independent of field of view size. ¿ American Geophysical Union 1996