The effects of core suture geometry on the mechanics of failure in flexor tendon surgery are investigated. Forty porcine flexor tendons were repaired using a Kessler; a Kessler-Pennington; a double Kessler; a continuous Kessler; and a cruciate repair. At maximum breaking strength, the cruciate repair gapped more then the double Kessler (12.8 mm vs 9.1 mm), but the double Kessler was less strong (37N vs 45 N). Transverse narrowing was 22% and 24% for the Kessler and the Kessler-Pennington, 11% for the double Kessler, and 0% for the continuous Kessler and the cruciate repair. Kessler-type sutures failed by suture breakage and the cruciate repair by pull-out. Under load, the transverse part of the Kessler sutures narrows, allowing longitudinal parts to lengthen, leading to gapping. The double Kessler shortened transverse segment decreases gapping. Eliminating a transverse component (the cruciate repair) decreased gapping, but the cruciate failed at higher loads by suture pull-out.