225 patients with a rupture of the anterior cruciate ligament were randomized to undergo reconstruction with either a double-stranded semitendinosus-gracilis graft using a two-incision technique (Group I), reconstruction with a patellar ligament graft using a two-incision technique (Group II), or reconstruction with a patellar ligament graft using a single-incision endoscopic technique (Group III). Patients were followed for a minimum of 6 years and assessed for the rate of graft failure, amount of instability, knee strength, radiographic signs of degenerative changes, and functional outcome. Group I had significantly fewer patients with quadriceps strength deficits compared to Group III, and Group III had significantly fewer than Group II. Hamstring strength deficits were significantly fewer in Group III compared to Group II, and Group II had significantly fewer than Group I. The groups were not significantly different in other outcome measures.