A brief description is given of how temporal and spatial variability in the earth's radiative behavior influences the goals of satellite radiation monitoring systems and how some previous systems have addressed the existing problems. Then, results of some simulations of radiation budget monitoring missions are presented. These studies led to the design of the Earth Radiation Budget Experiment (ERBE). A description is given of the temporal and spatial averaging algorithms developed for the ERBE data analysis. These algorithms are intended primarily to produce monthly averages of the net radiant exitance on regional, zonal, and global scales and to provide insight into the regional diurnal variability of radiative parameters such as albedo and long‐wave radiant exitance. The algorithms are applied to scanner and nonscanner data for up to three satellites. Modeling of daily shortwave albedo and radiant exitance with satellite sampling that is insufficient to fully account for changing meteorology is discussed in detail. Studies performed during the ERBE mission and software design are reviewed. These studies provide quantitative estimates of the effects of temporally sparse and biased sampling on inferred diurnal and regional radiative parameters. Other topics covered include long‐wave diurnal modeling, extraction of a regional monthly net clear‐sky radiation budget, the statistical significance of observed diurnal variability, quality control of the analysis, and proposals for validating the results of ERBE time and space averaging.