Accurate material characteristics of human upper extremity tendons are important for determining optimal surgical repair techniques and for developing 3-dimensional computer models. Determining the mechanical and viscoelastic material properties of these tissues are dependent on accurately measuring cross-sectional area, using a testing system which minimizes experimental error, and choosing an appropriate viscoelastic model. Statistical examination of the cross-sectional area variation along the mid-substance length revealed that while some tendon types examined did have a significant difference in cross-sectional area between measurement locations, others did not. Statistical comparisons between the fits of Fung's quasi-linear viscoelastic model and of Schapery's non-linear viscoelastic model to a single stress relaxation showed that Schapery's model was significantly better for only one tendon type and no significant difference between model fits was seen for all other tendon types. Finally, ultimate strength and strain, stiffness, Young's modulus, percent relaxation, and the viscoelastic constants for both the Fung and Schapery models were determined for a series of upper extremity human tendons. The mechanical and viscoelastic characteristics determined in this study are therefore more representative of healthy tissues compared to previous studies since fresh tissue was examined and more accurate and precise testing instruments were used.