A case control study of Scottish children aimed to identify risk factors for leukaemia and other cancers operating in the prenatal environment, during delivery and neonatally. Cases (0-14 years) were age-and sex- matched to two population-based controls and details abstracted from the mother's hospital obstetric notes. Analyses of 144 leukaemias (124 acute lymphoblastic leukaemias (ALL)), 45 lymphomas, 75 central nervous system (CNS) tumours and 126 'other solid tumours' were conducted using conditional logistic regression. The presence of a neonatal infection significantly reduced the risk of ALL (odds ratio (OR) 0.49, 95% confidence interval (CI) 0.26-0.95), particularly in 0- to 4-year-olds. Positive swab tests confirmed 47% of ALL cases with any infection and 46% of controls. This is consistent with the hypothesis that early exposure to infections may reduce the risk of childhood ALL. Asphyxia at birth significantly increased the risk of leukaemia, which was accounted for by ALL. For the 'other solid tumours' higher levels of maternal education were inversely associated with risk (OR 0.59, 95% CI 0.37-0.94) but positively associated with antibiotics (OR 2.16 95% CI 1.10-4.25) and respiratory tract infections (OR 14.1, 95% CI 1.76-113.7) in pregnancy. No obvious plausible patterns of risk were detected either within or across disease subgroups.