BACKGROUND: Colonization of the respiratory tract with UU has been associated with the development of CLD. However, culture positivity alone is a poor predictor of this outcome. Isolation of specific UU serotypes may help identify those babies who are truly at high risk and those that might benefit from antimicrobial therapy. OBJECTIVE: To determine the distribution of UU serotypes isolated from the respiratory secretions of preterm neonates and to assess the association between colonization with specific serotypes and the development of CLD. METHODS: As part of a randomized trial to evaluate Alpha-1 Antitrypsin supplementation for the prevention of CLD, endotracheal tube (ETT) aspirates taken on days 0,4 and 14 were cultured for UU. Isolates were serotyped by immunoperoxide assay. CLD in survivors was defined by the requirement for supplemental oxygen at 36 weeks' corrected gestational age. RESULTS: 106 babies with birth weights of 600-1250 g were enrolled. At least one ETT aspirate was obtained and cultured in 103 cases. UU was isolated from 30/91 (33%) babies surviving to 36 weeks and serotyping was successful in all but one. 29 patients yielded 51 isolates from the 14 known serotypes: 38 (75%) were from the parvum biovar (serotypes 1,3,6 and 14) and 30 (59%) were serotypes 3 or 14. 24/29 (83%) UU positive patients had at least one isolate from the parvum biovar. 26/91 (29%) babies developed CLD. There was a weak association between UU culture positivity and the development of CLD (RR=1.49, CI 0.78-2.84). Confining the analysis to either a specific serotype or biovar did not improve the predictive ability. CONCLUSIONS: UU isolates from the respiratory tracts of very preterm neonates in Toronto are predominantly from the parvum biovar. UU colonization appears to be weakly associated with the development of CLD. Serotyping does not aid in the identification of infants at increased risk for this condition.Funding received from Bayer/Canadian Red Cross Society Research and Development Fund