[en] A total of 7109 C 57 Black mice of either sex were either sham-, neutron- or cobalt-irradiated at the age of 33+-3 days. The dose range extended from 3.2 to 47.2 rads for neutrons and from 18 to 141 rads for γ rays. This experiment was aimed at assessing relative biological effectiveness (RBE) and its possible dose dependence from radiation mortality and morbidity data, using actuarial statistics. The primary goal was to define the spontaneous or basic tumour spectrum in control animals of either sex, which included a major proportion of lymphocytic lymphomas, followed by reticulum cell lymphomas. A variety of other tumours were observed, among them various adenocarcinomas, squamous cell carcinomas and leiomyosarcomas of the gastrointestinal tract, liver and spleen hemangiomas, micro- and macrofollicular hyperplasia as well as papillary adenomas of the thyroid, fibromas and fibrosarcomas of connective tissue, muscles and bones, myomas and fibrosarcomas of uterus, luteomas and cysts of the ovary, papillary and transitional cell carcinomas of kidneys and an occasional mammary adenocarcinoma or skin epithelioma. Modulation of the basic tumour pattern by neutrons or cobalt was primarily operative on the reticular tissue. Age-specific incidence rates for combined mature and immature lymphocytic lymphomas decreased following neutron or cobalt irradiation at higher dose levels. In reticulum cell sarcomas, however, the tumour type evidently shifted almost entirely from the mature to the immature type. Furthermore, all rates increased markedly following irradiation at all dose levels. One major finding in all groups was the non-linearity of tumour response. Furthermore, it appeared that the control as well as the radiation-modulated tumour spectra were likely to proceed from intercompetitive stochastic processes rather than from some probabilistic random distribution. Some implications of these primary results are discussed