Non-operative treatment of Achilles tendon ruptures is associated with an increased risk of rerupture. We hypothesized that this is due to inferior mechanical properties during an early phase of healing, and performed a randomized trial, using a new method to measure the mechanical properties. Tantalum markers were inserted in the tendon stumps, and tendon strain at different loadings was measured by stereo-radiography (Roentgen stereophotogrammetric analysis) at 3, 7 and 19 weeks and 18 months after injury. Thirty patients were randomized to operative or non-operative treatment. The primary out-come variable was an estimate for the modulus of elasticity at 7 weeks. Strain per force, cross-sectional area and tendon elongation were also measured. The functional outcome variable was the heel-raise index after 18 months. There was no difference in the mean modulus of elasticity or other mechanical or functional variables between operative and non-operative treatments at any time-point, but strain per force at 7 and 19 weeks had a significantly larger variation in the non-operative group. This group, therefore, might contain more outliers with poor healing. The modulus of elasticity at 7 weeks correlated with the heel-raise index after 18 months in both treatment groups (r2=0.75; P=0.0001). This correlation is an intriguing finding.