Background <p>Neonatal jaundice is a prevalent neonatal condition, with improper management potentially leading to severe neurological damage. Significant disparities in care quality exist globally, highlighting the urgent need for a standardized, disease-specific nursing quality evaluation system.</p> Aims and objectives <p>This study aimed to construct a scientifically rigorous and reliable sensitive indicator system for evaluating the quality of neonatal jaundice care, based on Donabedian's structure-process-outcome model.</p> Methods <p>A methodological study design was employed. An initial draft was developed through a systematic literature review and semi-structured interviews. Subsequently, 18 experts participated in two rounds of Delphi consultations to reach a consensus on the indicators.</p> Results <p>The effective response rates for the two consultation rounds were 90.0% and 100.0%, respectively. The expert authority coefficients were 0.851 and 0.879. The final system comprises 3 first-level, 10 second-level, and 24 third-level indicators.</p> Conclusion <p>The successfully constructed indicator system demonstrates strong scientific validity and reliability. It provides a consensus-based, standardized framework for assessing neonatal jaundice care quality, requiring further empirical testing before large-scale implementation.</p>