OBJECTIVES: This study compared the fatigue life of nailed proximal third tibial fractures stabilized with either three or four proximal screws using commercially available nails with both locked (through threaded holes or end caps) and nonlocked proximal interlocking screw configurations. METHODS: Eight paired and two independent tibiae of known bone mineral density were acquired, divided into three groups, and implanted with three different commercially available nails (n = 6/group). Nails were all 10 mm in diameter and individually sized for length. Individual tibiae from a given pair received different nails. Based on nail design, Nail A received four proximal screws (three that lock into the nail), whereas Nails B and C each received three proximal nonlocking screws. Standard end caps were used with all nails. As a result of its design, in Nail B, the most proximal interlocking screw was "locked" by the nail end cap. All nails used two distal screws. After implantation, an unstable proximal third fracture was created and specimens were tested with combined axial and torsional loads of 40 to 400 N and 0.11 to 1.1 Nm for 500,000 cycles or until failure. RESULTS: The fatigue life of Group A was significantly greater than either Groups B or C (P < 0.001 in both cases) with a mean cycle to failure of 392,977 versus 86,476 and 64,595 cycles for Nails B and C, respectively. Fatigue life of Group A was greater or equivalent to all contralateral tibiae; Group B outlasted all contralateral Group C limbs and the Group C constructs did not outlast any contralateral limbs. Bone mineral density correlated positively and significantly with fatigue life across all three groups (P < 0.001). CONCLUSIONS: In this study, proximal segment stability was improved with a greater quantity of screws and with locked interlocking screws.