PURPOSE: This study tests the idea of using radiodensity from computed tomography to quantitatively evaluate the healing of ruptured Achilles tendons. METHODS: The radiodensity of the healing tendons in sixty-five patients who were treated for Achilles tendon rupture was measured. The hypothesis was that density would correlate with an estimate for e-modulus, derived from strain, measured by radiostereometry with different mechanical loadings. RESULTS: Radiodensity 7 weeks after injury was decreased to 67 % (SD 11) of the contralateral, uninjured tendon. There was no improvement in radiodensity from 7 to 19 weeks, whereas at 1 year, it had increased to 106 % (SD 7). Only 2 of 52 measured values at 1 year were lower than the highest value at 19 weeks, i.e. there was minimal overlap. The variation in radiodensity could explain 80 % of the variation in e-modulus, but radiodensity correlated only weakly with e-modulus at each time point separately. At 1 year, both radiodensity and e-modulus correlated with functional results, although weakly. CONCLUSIONS: From 19 weeks onwards, radiodensity appears to reflect mechanical properties of the tendon and might to some extent predict the final outcome. Radiodensity at 7 weeks is difficult to interpret, probably because it reflects both callus and damaged tissues.