The relationship between frequency of consumption of whole grain food and risk of selected neoplasms has been analysed using data from an integrated series of case-control studies conducted in northern Italy between 1983 and 1996. The overall dataset included the following incident, histologically confirmed neoplasms: oral cavity and pharynx 181, oesophagus 316, stomach 745, colon 828, rectum 498, liver 428, gallbladder 60, pancreas 362, larynx 242, breast 3,412, endometrium 750, ovary 971, prostate 127, bladder 431, kidney 190, thyroid 208, Hodgkin's disease 80, non-Hodgkin's lymphomas 200, multiple myelomas 120. Controls were 7,990 patients admitted to hospital for acute, non-neoplastic conditions, unrelated to long-term modifications in diet and not likely to have been caused by tobacco or alcohol use. Odds ratios (OR) for subsequent scores (never/occasional/frequent) of whole grain food consumption were derived after allowance for age, sex, education, smoking, alcohol intake and body mass index. High intake of whole grain foods consistently reduced risk of neoplasm at all sites, except thyroid. The ORs for the highest category of consumption were 0.2-0.3 for upper digestive and respiratory tract neoplasms, 0.5 for stomach, colon and gallbladder, 0.7 for rectum, 0.6 for liver, 0.8 for pancreas and prostate, 0.9 for breast and endometrium, 0.6 for ovary, 0.4 for bladder and kidney, 1.3 for thyroid and around 0.5 for lymphomas and myeloma. The tests for trend in risks were significant for all neoplasms, except pancreas, endometrium, Hodgkin's disease and multiple myeloma. No significant heterogeneity was found across strata of age at diagnosis, sex, education, smoking habit, alcohol intake and body mass index. Thus, even in the absence of a univocal and satisfactory biological interpretation, the consistency of the patterns observed indicate that, in this population, higher frequency of whole grain food intake is an indicator of reduced risk of several neoplasms.