Referring to Lijinsky’s suggestion that lifetime carcinogenicity testing of drugs and environmental chemicals in rats might not be predictive of their carcinogenicity in longer living species such as man, Miller observed that all our present information on the natural history of cancer, both spontaneous and induced, in the experimental animal matches that of man. Lijinsky agreed that this was so, but pointed out that there are at present no good experimental data to indicate whether the latent period is related to dose of the carcinogen or to life span of the species. He suggested that the following experiment might be informative: administration of the same carcinogen to species with widely differing lifespans and determination of whether the tumors appear after a fixed time interval or at the same stage in the lifespan. In the ensuing discussion, Rentsch observed that rats reared under strictly germ-free conditions live appreciably longer (4–5 years) than rats maintained under the usual laboratory conditions (2–3 years), and suggested that this fact could be used experimentally to explore this problem. In reply, Lijinsky agreed but noted that the maintenance of appreciable numbers of rats under germ-free conditions would be very expensive.