Scoliosis is defined as the three-dimensional (3D) structural deformity of the spine with a radiological lateral Cobb angle (a measure of spinal curvature) of ≥10° that can be caused by congenital, developmental or degenerative problems. However, those cases whose etiology is still unknown, and affect healthy children and adolescents during growth, are the commonest form of spinal deformity, known as adolescent idiopathic scoliosis (AIS). In AIS management, early diagnosis and the accurate prediction of curve progression are most important because they can decrease negative long-term effects of AIS treatment, such as unnecessary bracing, frequent exposure to radiation, as well as saving the high costs of AIS treatment. Despite efforts made to identify a method or technique capable of predicting AIS progression, this challenge still remains unresolved. Genetics and epigenetics, and the application of machine learning and artificial intelligence technologies, open up new avenues to not only clarify AIS etiology, but to also identify potential biomarkers that can substantially improve the clinical management of these patients. This review presents the most relevant biomarkers to help explain the etiopathogenesis of AIS and provide new potential biomarkers to be validated in large clinical trials so they can be finally implemented into clinical settings.