In a study of 30 months duration, involving 600 male and 600 female Wistar rats fed on 12 different diets/dietary regimes, none of which involved deliberate exposure to any known genotoxic carcinogen, highly significant between-group differences were observed in survival and incidence of various neoplastic and non-neoplastic diseases. A full report of the findings is being prepared. Here we report that, irrespective of diet or dietary regime, there were highly significant correlations of body weight at 29 weeks of age with premature death (P less than 0.0001 in both males and females), with development of benign or malignant neoplasm of any site (P less than 0.0001 in males and P less than 0.01 in females) and with development of malignant neoplasm at any site (P less than 0.0001 for sexes combined). Numerous kinds of neoplasm contributed to these overall correlations. The most significant were pituitary tumour (P less than 0.0001), mammary gland tumour (P less than 0.0001), squamous or anaplastic carcinoma of the jaw (P less than 0.001), and subcutaneous mesodermal tumours (P less than 0.05). The 20% of rats that were heaviest at 29 weeks were more than twice as likely to die prematurely than the lightest 20% (2.56 times--males, and 2.11 times--females), and almost twice as likely to develop a malignant tumour (1.87 times for the sexes combined). These findings have important implications for the design and interpretation of carcinogenicity tests in rodents and of laboratory and human studies of relationships between diet, ageing-related degenerative diseases, and cancer.