Data on 538 mothers and their small-for-dates babies were analysed to ascertain whether any specific maternal factors were associated with increased perinatal risks. There were 34 deaths; 21 (62%) of these babies had major congenital abnormalities. Among the survivors perinatal morbidity was also markedly increased when the baby was abnormal. No direct associations were found between abnormal babies and any maternal factors. Few differences were found in the incidence of mortality and morbidity factors according to maternal height, weight, weight gain in pregnancy, social class and smoking habits. There was a significantly higher death rate when the mother was pre-eclamptic, and the incidence of seven other morbidity factors was also increased. The only adverse effects of maternal hypertension without pre-eclampsia was a higher instrumental delivery rate, and more of these babies were tube-fed. There were only significantly more perinatal deaths among multiparous women whose previous babies had been of average birthweight. This was due to an excess of congenitally abnormal babies in this group. Only 35% small-for-dates babies escaped any perinatal problems.