Finite element analysis (FEA). The aim of this study was to determine biomechanical differences between traditional growing rod (TGR) and spring distraction system (SDS) treatment of early-onset scoliosis. Many "growth-friendly" implants like the TGR show high rates of implant failure, spinal stiffening, and intervertebral disc (IVD) height loss. We developed the SDS, which employs continuous, dynamic forces to mitigate these limitations. The present FEA compares TGR and SDS implantation, followed by an 18-month growth period. Two representative, ligamentous, scoliotic FEA models were created for this study; one representing TGR and one representing SDS. initial implantation, and up to 18 months of physeal spinal growth were simulated. The SDS model was continuously distracted over this period; the TGR model included two additional distractions following index surgery. Outcomes included differences in rod stress, spinal morphology and iVD stress-shielding. Maximum postoperative von Mises stress was 249MPa for SDS, and 205MPa for TGR. During the 6-month TGR distraction, TGR rod stress increased over two-fold to a maximum stress of 417MPa, compared to a maximum of 262 MPa in the SDS model at 6-month follow-up. During subsequent follow-up periods, TGR rod stress remained consistently higher than stresses in the SDS model. Additional lengthenings in the TGR model led to a smaller residual curve (16.08) and higher T1-S1 growth (359 mm) at 18-month follow-up compared to the SDS model (26.98, 348 mm). During follow-up, there was less stress-shielding of the IVDs in the SDS model, compared to the TGR model. At 18-month follow-up, upper and lower IVD surfaces of the SDS model were loaded more in compression than their TGR counterparts (mean upper: +112 ± 19N; mean lower: +100 ± 17N). In the present FEA, TGR treatment resulted in slightly larger curve correction compared to SDS, at the expense of increased IVD stress-shielding and a higher risk of rod fractures. N/A.