The purposes of this study was 1) to assess the plantar flexion and dorsiflexion peak torques (PT) of the ankles at 30, 90 and 240.. s-1 in 101 patients (86 men, 15 women) who had been operated on for unilateral, complete closed Achilles tendon (AT) rupture at Oulu University Hospital, Oulu, Finland, in the period 1987-1992, 2) to determine whether comparison between the legs shows any impairment, 3) to study whether the weakness is speed-dependent, 4) to determine at what angular displacement of the ankle the weakness is manifested, 5) to study how the results correlate with age, type of operation and follow-up time. The Lido Multi-joint II dynamometer was used for the measurements. There was no significant dorsiflexion weakness detectable upon comparison between the legs, but the mean relative peak torque deficits in the injured limb were 8.4, 9.0 and 3.0% at 30, 90 and 240.. s-1 respectively for the men and 15.0, 16.6 and 6.4% for the women. The mean percentage torque differences were significantly greater in the women at all the test speeds (p < 0.05). The difference in PT was significantly greater at the two low test speeds (30 and 90.. s-1) than at the high speed (240.. s-1, p < 0.001). The weakness was manifested at an angular displacement of 80-120 degrees, where the average peak work (PW) difference between the two legs was significant in both sexes (p < 0.05). The patient's age (21-63), the type of operation (Lindholm or Silfverskiöld technique) and the follow-up time (0.7-6.7 years) did not significantly affect the results. In conclusion, and AT rupture implied an average 3.0 to 16.6% impairment in isokinetic plantar flexion muscle strength. The impairment was greater in the women, was manifested at an angular displacement of 80-120 degrees, and was greater at low test speeds. Age, type of operation and follow-up time did not account for the PT differences between the patients.