Interlocking nails have been available for the treatment of humeral shaft fractures in our unit since 1991. A retrospective review was carried out to assess the results of all patients receiving a nail over a 3-year period. Forty-three patients with fractures of the humeral shaft were treated by locked intramedullary nailing, using the Russell-Taylor nailing system. The indications for operation were pathological fractures (22 cases), multiple injuries (nine), failure of other treatment (four), or unstable isolated fractures (eight). There was one open fracture. Most nails were inserted in an antegrade direction from the proximal humerus. Notes and radiographs were reviewed on all pateints. Those with non-pathological fractures were recalled for clinical assessment including Neer shoulder score.Overall there were few technical problems with nail insertion. Additional fracture comminution occurred in five patients but affected stability in only one.Twenty-six nails were inserted into 22 patients with pathological fractures. The objective of fracture stabilization and pain relief for the remaining life-span of the patient was achieved in all but one patient, with few complications. Mean survival was 21.6 weeks (range, 3-64 weeks).Out of 21 non-pathological fractures, there were seven which had failed to unite 8 months after nailing (33 per cent), five of these being among the 17 cases nailed soon after injury (29 per cent). This suggests that non-union is a major problem. Four out of 15 patients reviewed clinically had poor or unsatisfactory shoulder function.These results do not suggest an advantage for interlocking nailing over plate fixation in non-pathological fractures, although more stable fixation may be obtained in the presence of osteoporosis. For pathological fractures the procedure provides reliable stabilization and pain relief.