Objective: Preterm infants are prone to increased bilirubin burden and display adverse outcomes if left unmonitored; therefore, predicting an increased bilirubin production is of paramount importance. Methods: We aimed to evaluate carboxyhemoglobin (COHb) levels in moderate (GA: 32(0/7)-33(6/7)) and late preterm (GA: 34(0/7)-36(6/7)) infants to assess whether this molecule could be used as an early predictor of phototherapy requirement. Results: A total of 221 infants were enrolled in the study. On admission, carboxyhemoglobin levels of infants who received phototherapy were significantly higher than that of infants who did not require this treatment, and this difference persisted in the consecutive hours (median (min-max): 1.2% (0.3-1.7) versus 0.8% (0.4-1.1); p < .001). The initial and consecutive COHb levels showed positive correlation (r = 0.77, p < .001). In the post-hoc analysis, direct antiglobulin test positivity significantly affected phototherapy requirement (p < .001). Receiver operating characteristics analysis showed that a COHb level of >= 0.95% was found to have a sensitivity of 90% and a specificity of 88%. Multinomial logistic regression analysis demonstrated that high COHb levels on admission significantly increased the likelihood of phototherapy requirement when adjusted for covariants (adjusted odds ratio: 2.2; 95% confidence interval: 1.4-3.5; p < .001). Conclusion: Carboxyhemoglobin measurement can be simply used to predict preterm infants who will require phototherapy.