BACKGROUND This study compared the fixation stability of two treatments for humeral shaft fractures with segmental bone loss during cyclic, physiologic loading. METHODS Six matched pairs of human humeri received either a 10-hole broad dynamic compression plate or a locked antegrade inserted humeral nail applied to a humeral diaphyseal osteotomy with a 1.5-cm gap defect. The bone-implant humeral constructs were axially loaded for 10,000 cycles at 250 N and 500 N, with measurements of gap displacement and calculation of construct stiffness. The specimens were then loaded to failure. RESULTS Cyclic loading showed no difference between the two groups for average gap displacement or construct stiffness. The intramedullary nail constructs failed by humeral shaft splitting (n = 4) or head cut-out (n = 2) at an average of 958.3 N, whereas the plate constructs failed by humeral shaft splitting and screw pull-out (n = 3) or plate bending (n = 3) at an average of 641.7 N (p < 0.001). CONCLUSIONS Although both methods offer similar fixation stability under physiologic loads, the higher load to failure demonstrated by intramedullary nail fixation may have implications for the patient with multiple injuries for whom partial weightbearing on the injured upper extremity may be necessary.