Previous studies have suggested that perinatal factors influence the risk for asthma but population studies on perinatal factors and risk for hay fever are few. We studied the effect of perinatal factors on the risk for hay fever among adolescent twins by a questionnaire study involving five consecutive nation-wide birth cohorts of 16-year-old twins and their parents. The risk for parent-reported, doctor-diagnosed hay fever in the adolescents associated with several perinatal characteristics was assessed with logistic regression analysis among individuals and by a discordant pair analysis. In the univariate analysis of the birth factors, the risk for hay fever increased with increasing birth weight (p for trend = 0.048, OR for those > or = 3000 g 1.35, 95% CI 0.91-2.02 compared to those < 2000 g) and gestational age (p for trend = 0.04, OR for those born after 40 weeks of gestation 2.24, 95% CI 1.03-4.86, compared to those born before 33 weeks of gestation) and was lower in those subjects hospitalised in the neonatal period (OR 0.74, 95% CI 0.58-0.93). Because of significant interactions between parental hay fever status and birth factors (ponderal index, p = 0.03 and maternal age p = 0.04), stratified analysis were performed. The positive association between birth weight and hay fever was most obvious among adolescents with no parental history of hay fever (p for trend = 0.03). Similar, though not significant, trends were found with other birth factors among these families, whereas no such trend was found among adolescents with parental hay fever, suggesting that gestational maturity increases the risk for hay fever in the absence of genetic predisposition. However, of the perinatal factors only neonatal hospitalisation (OR 0.75, 95% CI 0.59-0.96) remained a significant risk factor for the development of hay fever, when adjusted for non-perinatal factors.