Consistent and validated data sets of satellite-borne radiances and of a large variety of products describing the characteristics of terrestrial cloud and radiation fields have been produced within the International Satellite Cloud Climatology Project (ISCCP) covering the years 1983 through to 2003. A subset (annual and seasonal averages of the 5 year period 1991 to 1995) is used in this paper to discuss in greater detail the effect of clouds on the radiation fields at the upper and lower boundary of the atmosphere and in particular on the loss and gain (vertical divergence) of radiant energy by the atmosphere itself. Although this subset covers the effects of the Pinatubo eruption (June 1991) and of the strong El Nino event in 1992-93, which indeed caused 'anomalies in the average aerosol and cloud fields in the tropics and subtropics'. However, our regional averages of the radiation budget at the top of the atmosphere and at ground over a period of 5 years should be within 2-5 W m(-2) of longer term averages. We find very interesting spatial patterns in the global distributions of all quantities, which can be explained in part by various cloud field characteristics and by the continental surface characteristics. Most are known from similar studies with radiation budget measurements. Possibly for the first time, we show global fields of the vertical flux divergence of solar and terrestrial radiation within the atmosphere and of the effects of clouds. Both polar regions, various portions of China and the areas of persistent subtropical maritime stratocumulus fields over the Pacific and Atlantic Oceans and of cloud fields associated with the intertropical convergence zone (ITCZ) offer specific features for further analyses. This ISCCP data set seems to underestimate the absorption of solar radiation in the tropical and subtropical atmosphere by about 10 to 20 W m(-2). There is a disagreement of about 30 W m-2 in global averages of the gain and loss of solar and terrestrial radiation in the atmosphere between this and two other independent data sets, which needs thorough investigation, since such data are required to validate the radiation budgets within circulation and climate models and for other climate Studies. Such an assessment of radiation budget data is now under way within the auspices of the World Climate Research Programme. Copyright (c) 2005 Royal Meteorological Society.