Intact tendons adapt slowly to changes in mechanical loading, whereas in healing tendons, the effect of mechanical loading or its absence is dramatic. The longevity of the response to a single loading episode is, however, unknown. We hypothesized that the tissue has a "memory" of loading episodes, and that therefore short loadings are sufficient to elicit improved healing. The Achilles tendon of 70 female rats was transected and unloaded by tail suspension for 12 days (suspension started on day 2 after surgery). Each day, the rats were let down from suspension for short daily training episodes according to different regimes: 15 min cage activity, or treadmill running for 15, 30, 60 or 2x15 min. Rats with transected Achilles tendon and full-time cage activity served as controls. The results demonstrated that full-time cage activity increased the peak force over 3 times, compared to unloading. Short daily loading episodes (treadmill running) increased the peak force about half as much as full-time activity. Prolongation of treadmill running above 15 min, or dividing the daily training in 2 separate episodes, had minimal further effect. This mechanical stimulation increased the cross sectional area, but had no effect on the mechanical properties of the repair tissue. The findings indicate that once the tissue had received information from a certain loading type and level, this is "memorized" and leads to a response lasting many hours. This suggests that patients might be allowed early short loading episodes following e.g. an Achilles tendon rupture, for a better outcome.