<p>Congenital hypothyroidism (CH) can be detected early during thyroid-stimulating hormone (TSH) screening based on dried blood spot (DBS) samples. However, it falls short in differentiating between neonates with elevated TSH levels and those with CH, which leads to many neonates undergoing secondary diagnosis through venous blood sampling. Metabolomics was used to analyze metabolic alterations in neonates with CH, and identify DBS-based metabolite biomarkers for CH screening to reduce secondary diagnosis. Based on non-targeted metabolomics, the metabolic alterations in neonates with CH were analyzed in a discovery set and novel biomarkers were identified. The results of the discovery set revealed that the metabolic alterations in neonates with CH were primarily in amino acid and lipid metabolism, and identified novel metabolite biomarkers were thyroxine, 2-piperidinone, and PC (14:0/20:4). Then, these biomarkers were validated in a validation set, and the screening performance was still satisfactory. A rapid 3.5-min targeted method for three potential biomarkers was further developed and used to analyze a confirmation set. Analysis of the confirmation set re-validated the reliability of the biomarkers, and a biomarker combinational model equation and an appropriate cutoff value were defined. Each set of DBS samples included neonates with health, hyperthyrotropinemia, and CH. The novel metabolite biomarkers from DBS samples demonstrate significant potential for CH screening in neonates, effectively reducing the requirement associated with secondary diagnosis of mis-screened neonates.</p> Graphical Abstract <p></p>