IN 1940, Cook et al. 1 reported that 2 : 2′-azo-naphthalene (I) produces cancers of the liver in mice. The compound was also given to rats by subcutaneous injection, and with the food; but in this species no neoplastic changes were found in the livers. There is a brief note in the 24th Annual Report of the British Empire Cancer Campaign, 1947, p. 147, that a few small areas of cholangiectasis were observed in three out of eighteen rats receiving this compound by the oral route. No further details have been published, but Dr. A. H. M. Kirby has kindly informed us (see Kirby, Ph.D. thesis, London, 1949) that he obtained a number of foci of hepatoma in rats surviving more than 500 days. These rats were maintained on a balanced diet and were given the azonaphthalene orally. This behaviour is in marked contrast to 4-dimethylaminoazobenzene (II, X = H), which is a liver carcinogen in rats, but which is very much less active in mice. However, 4-dimethylaminoazobenzene is only effective as a liver carcinogen when tested in rats which are maintained on a diet that is low in protein and in riboflavin2, and the original experiments with 2 : 2′-azonaphthalene were completed before the importance of the diet had been established. It therefore seemed of interest to re-test 2 : 2′-azonaphthalene for carcinogenic activity in rats maintained on a suitably restricted diet, for direct comparison with 4-dimethylaminoazobenzene. At the same time, an opportunity has been taken to test other azo-compounds which are also of interest in the study of the relationship between chemical constitution and carcinogenic activity2.