Objective: The aim of this study was to determine the value of pro- and anti-inflammatory cytokine levels in both blood and tracheal aspirate (TA) samples that were obtained within 24 h after birth for predicting bronchopulmonary dysplasia (BPD) development in premature infants. Material and methods: Premature infants, who were born before 32 weeks of gestation, weighing less than 1,500 g, and admitted with respiratory distress between September 2009 and December 2010, were enrolled. Tracheal aspirate samples and serum were obtained from all infants on the first day of admittance for evaluation of pro- and anti-inflammatory cytokine levels using ELISA. Results: The study included 102 premature babies of whom 31 (30%) had BPD diagnosed in the follow-up. Mild, moderate and severe BPD was diagnosed in 10 (32%), 14 (45%) and seven (23%) infants, respectively. Both serum and TA sample pro- inflammatory cytokine (TNF-alpha, IL-1 beta, IL-6) levels were significantly higher, and anti-inflammatory cytokine (IL-10) levels were significantly lower in infants who developed BPD compared with those who had no BPD. No significant differences were detected in either serum or TA sample pro- and anti-inflammatory cytokine concentrations in preterm infants with BPD in terms of BPD severity. Cut-off values of both serum and TA sample pro- and anti-inflammatory cytokine concentrations for predicting BPD were also determined. Conclusion: It is suggested that higher serum and TA pro- inflammatory cytokine (TNF-alpha, IL-1 beta, IL-6) concentrations, along with lower anti-inflammatory cytokine (IL-10) concentrations, might be used for predicting the development of BPD in premature infants with respiratory distress at birth.