<b><i>Background:</i></b> Routine and timely determination of total serum bilirubin (TSB) remains a challenge in many resource-limited countries with substantial burden of severe neonatal hyperbilirubinemia. Limited evidence exists on the potential usefulness of transcutaneous bilirubin (TcB) to identify infants who may require phototherapy based on possible treatment criteria in such settings. <b><i>Objective:</i></b> To compare the number of infants requiring phototherapy across different TSB criteria and determine the predictive performance of TcB under each criterion. <b><i>Methods:</i></b> Infants with paired TcB and TSB measurements in a maternity hospital in Lagos, Nigeria, were assessed for phototherapy based on TSB criteria by the American Academy of Pediatrics (AAP), National Institute for Health and Clinical Excellence (NICE) UK, and World Health Organization (WHO), and an absolute threshold of ≥12 mg/dL. The predictive performance of TcB across treatment criteria was evaluated with receiver operating curve analysis. <b><i>Results:</i></b> A total of 1,011 infants with a mean TcB of 10.54 ± 3.19 (range: 2.7-19.9) mg/dL and TSB of 9.63 ± 2.61 (range: 0.3-19.5) mg/dL were assessed. Some 60 (5.9%) infants required phototherapy by 1 or more TSB criteria, with TSB ≥12 mg/dL identifying 55 (91.7%) and AAP 27 (45%) of these infants. All infants identified by the NICE and WHO criteria were equally detected by the AAP criterion. TcB showed negative predictive values of 99.0-99.9%, and positive predictive values of 7.7-15.5% across all criteria. <b><i>Conclusions:</i></b> The number of infants requiring phototherapy varies significantly across treatment criteria. TcB may be useful in identifying infants who do not require phototherapy, but may also identify a high proportion of false positives that is burdensome in resource-limited settings.