This population-based register study examined if factors during the fetal and neonatal period influence the risk for the child to develop bronchial asthma (asthma). From the Swedish Hospital Discharge Register we identified children, born between 1987 and 1999, who had been hospitalized for asthma up to 2001. Thus, the outcome measure contains only hospitalized cases, not all children with asthma. Children younger than 2 yr at admission were excluded because of the uncertainty about the diagnosis of asthma in younger children. The remaining 14,803 children were compared with all children born the same years, recorded in the Swedish Medical Birth Registry, for information on pre- and perinatal characteristics. Odds ratios (ORs) were calculated with Mantel-Haenszel technique and 95% confidence intervals (CIs) with Miettinen's test-based method. The presence of various maternal and neonatal confounders were identified and adjusted for in the analyses. The association between some known factors and childhood asthma were confirmed: young maternal age, maternal smoking, period of unwanted childlessness, low maternal level of education, maternal diabetes, preterm birth, low birth weight, small-for-gestational age, caesarean section, and instrumental vaginal delivery. A number of neonatal characteristics were shown to be independent risk factors: sepsis or pneumonia, neonatal respiratory problems and treatments, neonatal icterus, and/or neonatal phototherapy. The association with icterus and phototherapy remained after exclusion of cases showing other neonatal risk factors and after adjustment for maternal factors (OR 1.27, 95% CI: 1.08-1.50), and increased to 1.5 if the children had been hospitalized for asthma more than once. In conclusion, our results suggest an association between neonatal icterus and/or treatment with neonatal phototherapy and hospitalized childhood asthma. This association needs further exploration.