BackgroundAdditional auxiliary rods have been used in spinal instrumentation across pedicle subtraction osteotomy to reduce stresses in the primary rods. The auxiliary rods can be connected through dual-rod-screws, fixed-angle multi-rod connectors or variable-angle multi-rod connectors. The objective was to assess rod bending in conventional bilateral-rod construct vs. constructs with auxiliary rods. MethodsComputer models of two adult patients were developed to evaluate bending loads across a pedicle subtraction site in a control construct with bilateral rods vs. constructs with auxiliary rods bilaterally or unilaterally connected to the primary rods through either dual-rod-screws, fixed-angle multi-rod connectors, or variable-angle multi-rod connectors. Postoperative rod bending loads were computed and compared. FindingsNormalizing loads on the primary rods in the multi-rod constructs to the control construct, primary rod loads in multi-rod constructs were 17% to 48% lower than the control construct. Constructs with bilateral auxiliary rods through dual-rod-screws, fixed-angle multi-rod connectors, or variable-angle multi-rod connectors could result in similar primary rod bending loads. Bending loads on the auxiliary rods were higher or lower than those on the primary rods depending on how their curvatures matched the primary rods, and how they were locked onto the primary rods. InterpretationAuxiliary rods noticeably reduced the bending loads on the primary rods compared with a standard bilateral-rod construct. Loads in the auxiliary rods were higher or lower than those in the primary rods depending on how their curvatures matched the primary rods, and how they were locked onto the primary rods.