We investigated colonization with Ureaplasma urealyticum (Uu) in infants <30 weeks gestation and assessed the relationship to other risk factors influencing respiratory morbidity, plus the effect of treatment with erythromycin. Ventilated preterm infants [n = 155; median GA 26 (23-29) weeks] were cultured for Uu in endotracheal aspirate and nasopharynx. Colonized infants were randomly assigned to treatment with erythromycin 40 mg/kg/d, intravenously or orally. The rate of colonization was 29/155 (19%) and the Uu-colonized infants had lower mean gestational ages than the culture-negative infants (25 vs 26 weeks). For the colonized infants PROM (48% vs 12%), chorioamnionitis in the mother (46% vs 17%) and vaginal delivery (71% vs 29%) were more common. More colonized infants needed supplemental oxygen at 36 weeks' postconceptual age (p < 0.05). Erythromycin treatment was effective in reducing colonization with negative control cultures in 12/14 (86%) of treated infants. No significant differences were found between the colonized treated infants (n = 14) and those not treated (n = 14) in time with supplemental oxygen. Oxygen requirement at 36 weeks was related to lower gestational age, late appearance of PDA, late onset sepsis and signs of chorioamnionitis in the mother. We conclude that the Uu colonization is related to increasing immaturity, the presence of prolonged rupture of membranes, signs of chorioamnionitis and vaginal delivery. Treatment with erythromycin reduced colonization but did not significantly alter length of time with supplemental oxygen.