B-mode ultrasound is an established imaging modality for evaluating canine tendon injury. However, full extent of tendon injury often remains difficult to estimate, as small changes in sonographic appearance are associated with large changes in biomechanical strength. The acoustoelastic strain gauge (ASG) is an ultrasound-based tissue evaluation technique that relates the change in echo intensity observed during relaxation or stretching of tendons to the tissue's mechanical properties. This technique deduces stiffness gradient (the rate of change of normalized stiffness as a function of tissue strain) by analyzing the ultrasound dynamic images captured from gradually deforming tissue. ASG has been proven to accurately model strain and stiffness within tendons in vitro. To determine the feasibility and repeatability for in vivo ASG measurements of canine tendon function, stiffness gradients for the gastrocnemius tendons of 10 clinically normal dogs were recorded by two nonindependent observers at three sites (musculotendinous junction, mid tendon, and insertion). Average stiffness gradient indices (0.0132, 0.0141, 0.0136) and dispersion values (0.0053, 0.0054, 0.0057) for each site, respectively, were consistent with published mechanical properties for normal canine tendon. Mean differences of the average stiffness gradient index and dispersion value between observers and between limbs for each site were less than 16%. Using interclass coefficients (ICC), intra-observer (ICC 0.79–0.98) and interobserver (ICC 0.77–0.95) reproducibility was good to excellent. Right and left limb values were symmetric (ICC 0.74–0.92). Findings from this study indicated that ASG is a feasible and repeatable technique for measuring stiffness gradients in canine tendons.