The current nominal risk coefficients for radiation induced cancer are predominantly based on the follow-up of the atomic bomb survivors, i.e. a collective of persons exposed to a short duration, high dose rate exposure. It is one of the central issues in radiation protection, whether these data are representative also for protracted, low dose rate exposures. The International Commission for Radiological Protection has postulated a dose and dose rate reduction factor in the derivation of the nominal risk coefficients; but this factor has been debated. Direct observations on populations that were subjected to prolonged low dose exposures are, therefore, of special interest, and a major contribution to the problem may result from the long term observation of the population exposed at the beginning of the 1950s as a result of massive releases of fission products by a plutonium producing facility into the Techa river in the southern Urals. The present article deals with the cancer mortality in this population during the period 1950-1982.