Flexor tendon injuries are a common problem requiring sutured repair followed by early postoperative mobilisation. Results are better if active rather than passive flexion is allowed, however active movement increases the incidence of dehiscence. This has prompted examination of thermal bonding as a strategy for strengthening tendon repairs. Three experiments are described, using method and experimental design similar to those previously described. The strength of thermal bonds created between sections of bovine Achilles tendon at a range of temperatures and times was measured. Bonds were created by heating compressed sections in a water bath. Bond strength was measured on a tensiometer as load (N) and stress (N/mm2 and N m kg-1) then calculated using measured cross-sectional area, weight and length. Results were in broad agreement with those previously described, showing peak bond strength in samples heated to 62 degrees C for 8 min (median 17.11 N/cm2, SD 2.13). The effects of cooling to 0 degree C for 10 min immediately after bonding was to increase median bond strength when compared with non-cooled samples bonded under similar conditions (non-cooled, 9.11 N/cm2, SD 3.11; cooled, 11.96 N/cm2, SD 2.93). Statistical analysis using a quadratic regression model showed a close correlation between experimental data and curves modelled to best fit the data. Peak values for thermal bonds created using the methods described approached and in some cases exceeded that reported for the epitendinous element of modified Kessler tendon repairs in human digital flexor tendons. If such bonds could be created in vivo, thermal bonding may have a role in augmenting the strength of tendon repairs. The methods described are impractical for a non-laboratory environment. A laser would seem the most attractive alternative source of heat energy.