We compared the effects of reamed versus unreamed locked intramedullary nailing on cortical bone blood flow in a fractured sheep tibia model. A standardized spiral fracture was created by three-point bending with torsion, and each tibia was stabilized by insertion of a locked intramedullary nail. Eleven animals were randomized into two groups: one that had reaming before nail insertion and one that did not. Blood flow was measured in real time using laser Doppler flowmetry. Cortical bone perfusion measurements were made at three locations (proximal diaphysis, fracture site, distal diaphysis) and at eight time intervals (prefracture, postfracture, postreaming, postnail insertion, postlocking, and at 2-, 6-, and 12-week follow-up). All animals were killed at 12 weeks postoperatively. After reamed nail insertion, cortical bone perfusion was significantly decreased (p < 0.0009). After unreamed nail insertion, perfusion was decreased less (p < 0.003). Insertion of locking screws did not affect blood flow. Cortical bone perfusion was greater in the unreamed group at completion of the procedure (p < 0.011), at 2-week follow-up (p < 0.006) and at 6-week follow-up (p < 0.027). The findings suggest that cortical revascularization had occurred by 6 weeks in the unreamed group but not until 12 weeks in the reamed group. The study demonstrates that cortical circulation is spared to a greater degree by unreamed nailing. This may be advantageous in severe open tibial fractures where blood supply is already significantly compromised.