In order to place cloud interpretation from the TIROS satellite pictures on a more objective basis, methods have been developed for computing radiance and albedos from the pictures. Provision was made for multiple scattering in a Rayleigh atmosphere and absorption due to 0.28 cm of ozone. Effects due to Mie scattering were not included. Corrections have been determined for variations at the satellite across the filter-lens-videcon system, the shutter speeds and warm-up of the TV system and for variations at the ground across the cathode-ray tube-camera system. The photometry on test cases made use of rerun video tapes and controlled film sensitometry. Simultaneous with the times of satellite pictures, other pictures of the same clouds were obtained from a U-2 aircraft. Radiance and albedos of the reflecting surfaces in these pictures was determined in a manner similar to that from the satellite pictures. A comparison of albedos for the respective spectral response of the systems shows good agreement, although the values at high radiance levels appear more realistic when measured by the satellite. The standard deviation of a series of readings, from successive pictures, on the same reflecting cloud surface averaged about 15 per cent of the radiance. From this series of pictures, variations in albedo that might be due to non-Lambertian reflecting surfaces and changes in nadir viewing angle could not be isolated from the “noise” of the system. A summary of average albedos, determined from TIROS, for cloud types and terrestrial surfaces is presented. Typical albedos range from 32 to 36 per cent for cirrus and cirrostratus over land and from about 18 to 92 per cent for giant cumulonimbus. The application to TIROS photography of similar corrections and techniques will permit digital computation of albedo in “real time” and its presentation in the form of rectified mosaics. These digital data are also suitable for insertion into computer forecasts, compilation of climatological data and other special applications.