Knee function was evaluated by knee score, activity level, clinical findings and performance tests, muscle size by computerized tomography (CT), morphology by light (LM) and electron microscopy (EM), muscle function by electromyography (EMG) and isokinetic performance in 29 patients with chronic anterior cruciate ligament (ACL) tear. Preoperatively CT disclosed a significant mean atrophy of the quadriceps and nonsignificant changes of the other muscle areas of the injured leg. Morphology of m vastus lateralis of the injured leg was normal in more than half of the biopsies preoperatively, the rest showed signs of nonoptimal activation. Significant decreases in all isokinetic parameters were noticed together with significantly decreased EMG of the quadriceps muscle of the injured leg. Âfter surgical reconstruction the knees were immobilized in a cast for 6 weeks at either 30° or 70° of knee flexion. After cast removal CT showed significant decreases of all areas which also remained after training. The 30° group showed larger fibres (intracellular oedema) and more frequent morphological abnormalities than the 70° group. Fourteen weeks postoperatively the patients were allocated to either a combination of isometric and progressive resistance training or isokinetic training for 6 weeks. CT showed slightly larger areas at 20 weeks postoperatively than at 6 weeks. Morphological abnormalities were still prominent at 20 weeks postoperatively. Maximum isokinetic knee extensor mechanical output and endurance were markedly decreased at 14 weeks postoperatively but both improved progressively during the one year rehabilitation, mostly during the intensive 6 week training period but irrespective of training programme used. Fatiguability/endurance level improved over the preoperative level. Muscular work/integrated EMG was stable while EMG/t increased indicating neuromuscular relearning. The clinical result at 28 months foliowup was excellent or good in 93% of the patients and clinical stability improved in 66%. Independent upon primary knee immobilization angle or training programmes no differences could be demonstrated with respect to stability, range of motion, function or isokinetic mechanical output. Isokinetic performance was still significantly lower in the injured compared to the noninjured leg and not significantly different from the preoperative values. Morphology, only 6 cases, showed abnormalities similar to preoperative findings. In conclusion, the reason for the decreased maximum and total knee extensor performance in these patients with ACL tears is suggested to be nonoptimal activation of normal functioning muscle fibres depending on changes in knee joint receptor afferent inflow. No differences concerning the markedly improved postoperative clinical result could be seen between the different treatment modalities used. A nonoptimal muscular activation might explain the still decreased isokinetic performance present at followup. / <p>S. 1-40: sammanfattning, s. 43-137: 5 uppsatser</p> / digitalisering@umu.se