We compared the therapeutic results of two methods of exercise rehabilitation for achilles tendonitis: the “eccentric ankle drop” exercise, and the universal gym “concentric plantar flexion/dorsiflexion” exercise. Athletic patients (n = 17) with chronic achilles tendonitis were assigned randomly to one of the two treatment groups for 12 weeks. Patients were tested at 0, 4, 8, and 12 weeks for (a) average and peak torque for plantar-flexor concentric and eccentric muscle contractions (using the KIN/COM Dynamometer at 30 and 50°/s); and (b) a subjective evaluation of pain and return to activity rated on a scale from zero to 10. The eccentric protocol showed larger increases in the average and peak torque value compared with the concentric protocol, but these differences were not statistically significant. The plantar-flexion torque values increased for both test velocities in both groups over the 12-week course of treatment, but no treatment effect was observed between the two conditions. The eccentric protocol group had a significantly (p < 0.01) greater decrease in pain ratings averaged over the four testing sessions, and produced three times as many pain-free subjects at the end of the program. No significant differences were observed between the eccentric and concentric protocol groups with respect to the return-to-activity measures. We found that the subjective pain measure favors the use of eccentric exercise for the rehabilitation of achilles tendonitis.