A retrospective analysis of obstetric factors influencing mortality and morbidity of very premature infants (1500 g, less than or equal to 32 weeks' gestation) was undertaken. The study included 275 such infants born in the Department of Obstetrics of the University of Tübingen during the period January 1977 to June 1987. The caesarean section rate of very preterm infants increased from 28% during the period 1977-1982 to 87% during the period 1982-1987 (P less than 0.005), accompanied by an increase in survival rate from 63% to 70%. The improvement in survival rate was statistically significant for the group with birth weight 751-1000 g (P less than 0.01). The overall mortality rate was 31% after caesarean section and 36% after vaginal delivery. Amongst the causes of death of the non-survivors, acidosis was more frequent and amniotic infection syndrome less frequent in the infants delivered vaginally than in those delivered abdominally. The proportion of children with normal development at two years of age was significantly (P less than 0.02) greater amongst those born in 1982-1987 than in those born in 1977-1981. The interpretation of these findings is by no means clear but must include the hypothesis that the increased caesarean section rate may be incidental and in no way related to the improved outcome.