The flux density measured at satellite altitude with a fixed field of view radiometer differs from the true flux density reflected by the earth-atmosphere system within the field of view of the radiometer. This difference is due to angular response characteristics of the radiometer, solid angle effects due to geometry, and angular reflectance effects of the earth-atmosphere system. All of these effects lead to uncertainties in the interpretation of instantaneous earth radiation budget measurements. The differences between the true flux density and the measured flux density are shown to be significant when the field of view of the radiometer is large and when the atmosphere has a nonuniform, or spatially dependent, reflectance (albedo). A simulation experiment is described whereby the scene within the field of view of a nadir looking sensor is divided into a large number of equal area elements, each of which reflects radiation with one of two different reflectance models (corresponding to cloud-free and cloudy areas). The conditional mean values of the measured flux density, given values of the true flux density, are shown to differ significantly from the conditional means of the inverse problem, that of finding the mean value of the true flux density given a value for the measured flux density. The differences between the true flux density and the measured flux density are examined as a function of satellite altitude, field of view of the radiometer and solar zenith angle (including the effects of a terminator within the field of view) for both Lambertian and non-Lambertian reflectance models.