The inherent disadvantages of nerve grafting have made it necessary to find alternative techniques for treating segmental nerve loss. This study compares the techniques of nerve grafting and nerve expansion/repair for the management of nerve injuries with segmental nerve loss in an animal model. Bilateral segmental sciatic nerve defects were created in eight dogs. On one side a 2-cm segment was excised and replaced with a nerve graft; on the other side a 2-cm defect was created with ligatures and the nerve underwent preliminary expansion and then repair. Eighteen months later nerve conduction velocity (NCV), gastrocnemius contraction force (GCF), and muscle weight (GMW) were determined for the seven surviving animals. NCV for the expanded repair was 58.66 +/- 34.18 m/sec and 47.73 +/- 7.93 for the grafted repair (p = 0.4); GCF was 619.04 +/- 353.70 gm for the expanded repair and 726.80 +/- 415.78 gm (p = 0.2) for the grafted repair; and GMW was 82.80 +/- 5.68 gm for the expanded repair and 109.55 +/- 20.63 gm (p = 0.02) for the grafted repair. The data suggest that: 1) conventional tissue expansion techniques can be used successfully to repair segmental peripheral nerve defects; 2) NCV and GCF are comparable for grafting and expansion/repair techniques although GMW is significantly higher in the grafted group; and 3) nerve expansion/repair may prove to be a useful alternative to grafting.