In 1980-1983 members of IRESCC interviewed parents of 555 children with newly diagnosed cancer on topics of possible aetiological significance. Identical questions were asked of the parents of 1100 control children chosen from hospital admissions and general practitioner lists. Medical information was confirmed whenever possible by cross-checking with NHS records. Data for the 41 children with germ cell tumours and their 82 controls are reported here. The cases had more major congenital malformations than controls, including one neural tube defect. More case than control mothers and fathers reported occupational exposure to chemicals. Nine close relatives of cases had multiple primary tumours, which were often benign or of low-grade malignancy, compared with 1 hospital control and 3 general practitioner control relatives. Cases and controls differed with respect to birth weight and paternal age. No case-control differences were shown for: birth rank, maternal age, chronic illnesses and smoking, mothers' reproductive histories and oral contraceptive usage. In index pregnancies there were no case-control differences for maternal illness, infections, alcohol intake and X-ray and ultrasound exposure. There was no difference between cases and controls for the frequency of twinning the families.