Primary congenital hypothyroidism (CH) is the most common endocrine disease in children and one of the preventable causes of both cognitive and motor deficits. We present a genetic and bioinformatics investigation of rational clinical design in 17 Argentine patients suspected of CH due to thyroid dyshormonogenesis (TDH). Next-Generation Sequencing approach was used to identify variants in Thyroid Peroxidase (TPO) and Dual Oxidase 2 (DUOX2) genes. A custom panel targeting 7 genes associated with TDH [(TPO), Iodothyrosine Deiodinase I (IYD), Solute Carrier Family 26 Member 4 (SLC26A4), Thyroglobulin (TG), DUOX2, Dual Oxidase Maturation Factor 2 (DUOXA2), Solute Carrier Family 5 Member 5 (SLC5A5)] and 4 associated with thyroid dysembryogenesis [PAX8, FOXE1, NKX2-1, Thyroid Stimulating Hormone Receptor (TSHR)] has been designed. Additionally, bioinformatic analysis and structural modeling were carried out to predict the disease-causing potential variants. Four novel variants have been identified, two in TPO: c.2749-2 A > C and c.2752_2753delAG, [p.Ser918Cysfs*62] and two variants in DUOX2 gene: c.425 C > G [p.Pro142Arg] and c.2695delC [p.Gln899Serfs*21]. Eighteen identified TPO, DUOX2 and IYD variants were previously described. We identified potentially pahogenic biallelic variants in TPO and DUOX2 in 7 and 2 patients, respectively. We also detected a potentially pathogenic monoallelic variant in TPO and DUOX2 in 7 and 1 patients respectively. 22 variants have been identified associated with TDH. All described novel mutations occur in domains important for protein structure and function, predicting the TDH phenotype.