We review the data on the spectrum and isotropy of the microwave background radiation and the astrophysical processes that may produce spectral distortions and anisotropies. As yet no fully satisfactory explanation has been found for the submillimeter excess observed by Matsumoto et al. (1988). The most precise data at lambda > 1 mm disagree with nonrelativistic comptonization models which match the excess. Distortions produced by a very hot intergalactic medium yielding the X-ray background do not fit the submillimeter data. Very special requirements must be met for the interpretation in terms of high-redshift dust emission to work. Reported anisotropies on scales of several degrees and of tens of arcsec may be produced, at least in part, by discrete sources. Because the best experiments at cm wavelengths are close to the confusion limit, they provide interesting information on the large-scale distribution of radio sources.
