Background and Purpose: The use of locked nailing for the treatment of humeral nonunions is threatened by the potential of initial instability which may result in residual fracture gap and motion and is detrimental to fracture healing. This study investigated the biomechanical and clinical effects of adding interfragmentary wiring to reduce the initial instability in the treatment of humeral nonunion. Methods: The biomechanical study compared the mechanical properties of 3 fixation methods: locked nail alone, with staple, and with interfragmentary wiring. Composite humeri were transected to simulate nonunion and were retrograde nailed. Nail-bone constructs were examined by non-destructive bending tests and destructive torsional tests. The initial instability, fixation stiffness and strength were measured and compared. The clinical study compared 21 nonunions treated by locked nailing alone and 30 nonunions treated by locked nailing with interfragmentary wiring. The indication for wiring was persistent nonunion gap or motion during operation. The union rate, time to union, functional recovery of the adjacent joints and the time for functional recovery were compared between the treatment groups. Results: Biomechanically, in both bending and torsional tests, locked nailing with a staple or wiring resulted in significantly less initial instability than locked nailing alone. Test results for bending stiffness, torsional stiffness, or torsional strength did not differ significantly among the 3 fixation methods. In the clinical study, interfragmentary wiring had the advantage of providing tight nonunion compression, which was beneficial for nonunion healing and promoted early postoperative exercise. Patients treated with locked nailing with interfragmentary wiring had significantly shorter time to union and faster shoulder functional recovery (p<0.05). Conclusions: Adding interfragmentary wiring was an effective and safe procedure in humeral nonunions treated with locked nailing.