A population-based case-control study of incident primary malignant brain tumours diagnosed during 1985–1989 in children aged 0 to 14 years was carried out in the coastal conurbation of New South Wales comprising Sydney, Wollon-gong and Newcastle in the period 1988 to 1990. Personal interviews were conducted using a structured questionnaire with mothers of 82 cases and 164 control children individually matched to the cases by sex and age. Among the hypotheses examined were those related to: N-nitroso compounds (sources included diet, dummies, medications, tobacco smoke); factors associated with the birth of the child; trauma to the head; and irradiation (X-rays and electromagnetic radiation through electric blankets or water beds). Reported ever-use of a dummy increased the risk of childhood brain tumours (OR = 2.9, 95% CI 1.6 to 5.4), although there did not appear to be any consistent indication of rising risk with reported increased levels of use. Compared with children who had never used a dummy, categories of use during the first year of life of a maximum of “no more than I hour per day or night”, “several hours per day or night”, and “most of the day or night” had statistically significant odds ratios of 2.6, 3.4, and 2.7 respectively. Consumption of fruit by the child before the age of one appeared to be protective. No association was found between childhood brain tumours and birth weight, being the first-born child, or factors linked with the child's birth; head injuries; exposure to X-rays; contact with horses, or living on a farm; pesticide treatment of the house during the child's lifetime; or exposure to burning incense. © 1994 Wiley-Liss, Inc.