OBJECTIVE--To investigate if psychological distress during pregnancy is associated with increased risk of preterm delivery. DESIGN--Prospective, population based, follow up study with repeated measures of psychological distress (general health questionnaire), based on the use of questionnaires. SETTING--Antenatal care clinic and delivery ward, Aarhus University Hospital, Denmark. SUBJECTS--8719 women with singleton pregnancies attending antenatal care for the initial visit between 1 August 1989 and 30 September 1991; 5872 women (67%) completed all questionnaires. MAIN OUTCOME MEASURE--Preterm delivery. Estimation of gestational age at delivery was mainly based on early ultrasound measurements. RESULTS--In 197 cases (3.6%) the woman delivered prematurely (less than 259 days). A dose-response relation between psychological distress in the 30th week of pregnancy and risk of preterm delivery was found, but distress measured in the 16th week was not related to preterm delivery. Control of confounding was secured by the use of multivariate logistic regression models. Relative risk for preterm delivery was 1.22 (95% confidence interval 0.84 to 1.79) for moderate distress and 1.75 (1.20 to 2.54) for high distress in comparison to low distress. CONCLUSIONS--Psychological distress later in pregnancy is associated with an increased risk of preterm delivery. Future interventional studies should focus on ways of lowering psychological distress in late pregnancy.