The purpose of the current study was to test in vitro a new shape memory alloy suture for flexor tendon repair. Forty fresh-frozen human anatomic flexor superficialis and profundus tendons were divided and repaired via the cruciate four-strand technique using one of two suture materials (the shape memory alloy suture and the 4-0 Ethibond suture). The forces required to cause a 1, 2, and 3 mm gap, ultimate load to failure, and repair stiffness were compared. Twenty specimens of each suture material also were tensile tested for load to failure, tensile strength, and elongation at failure. The shape memory alloy suture had a significantly higher mean resistance force to 1, 2, and 3 mm gap formation than the 4-0 Ethibond suture (47 N versus 31 N, 51 N versus 36 N, and 57 N versus 41 N, respectively). The shape memory alloy suture repair was 40% stronger than the 4-0 Ethibond suture (61.9 +/- 8.8 N versus 44.3 +/- 10.6 N). Repair with the shape memory alloy suture was significantly stiffer than repair with the 4-0 Ethibond suture (8.1 +/- 1.0 N/mm versus 6.1 +/- 0.9 N/mm). The load to failure and tensile strength of the shape memory alloy suture were significantly higher than that of the 4-0 Ethibond suture. The values of elongation for the two materials were not significantly different. The results of the current study suggest that the shape memory alloy suture may be superior to the 4-0 Ethibond suture in resisting gap formation in the range of forces generated in the early rehabilitation protocol and may be the future material of choice for tendon repairs.