For examination of metabolic interrelationships in carcinogenesis between N-nitroso-2,6-dimethylmorpholine, N-nitrosobis(2-oxopropyl)amine (CAS: 60599-38-4), N-nitrosobis(2-hydroxypropyl)amine (CAS: 53609-64-6), and N-nitroso(2-hydroxypropyl) (2-oxopropyl)amine, each was given to a separate group of 20 female Syrian golden hamsters by gavage. All four compounds induced tumors of the pancreatic duct and lung tumors, but the incidences varied from one compound to another. In addition, N-nitrosobis(2-oxopropyl)amine and N-nitroso(2-hydroxypropyl)(2-oxopropyl)amine induced many hepatocellular and cholangiocellular neoplasms, which the other two compounds did not. On the basis of short time to death with tumors and the relatively low total dose administered, N-nitrosobis(2-oxopropyl) amine appeared to be the most potent carcinogen in the hamster among the four. N-nitroso(2-hydroxypropyl)(2-oxopropyl)amine was next in potency but was considerably weaker than N-nitrosobis(2-oxopropyl)amine. N-Nitroso-2,6-dimethylmorpholine, which was similar in potency to N-nitroso(2-hydroxypropyl) (2-oxopropyl)amine, however, did not induce a significant incidence of liver tumors of any type; and N-nitrosobis(2-hydroxypropyl) amine was considerably less potent than the other three compounds. These results did not support the opinion of N-nitroso(2-hydroxypropyl) (2-oxopropyl)amine as the proximate carcinogenic metabolite of all three compounds in the Syrian hamster but instead suggested that these compounds might have acted through formation of different and yet unknown carcinogenic intermediates.