The purpose of this study was (a) to evaluate by ultrasonography the healing of the patellar tendon after its mid‐third was removed for anterior cruciate ligament (ACL) reconstruction in two randomized groups of patients in whom the tendon donor site was either left open or closed: (b) to compare clinical, radiographic, and isokinetic studies of these two groups to evaluate the incidence of patellofemoral disorders. We performed 61 ACL reconstructions (22 males, 39 females) using the arthroscopically assisted in‐out technique. All operations were performed by the same surgeon, and the patients were all subjected to the same postoperative protocol. The tendon defect was left open in 25 subjects (group A) and was closed in 36 subjects (group B). Postoperative patellar tendon behavior was evaluated in these two groups by ultrasonography at 3, 6, 9, and 12 months. The vertical position of the patella was measured in the follow‐up lateral view at 45° of flexion and compared to that of the untreated knee. A clinical evaluation was performed throughout the follow‐up period. and patellofemoral problems (pain, stiffness, patellofemoral crepitus) were evaluated and recorded using a modified Larsen and Lauridsen rating scale. Isokinetic evaluation was carried out at 6 months, and a quadriceps index of the two groups was recorded. Ultrasonography showed that healing of the patellar tendon initially progressed with a compensatory hypertrophy in width and thickness. The width was greater in group B (P<0.01). In group A we observed in the cross‐sections a characteristic image of two cords separated by a low signal bridge which we defined as a “binocular pattern”. Areas of high ultrasound signal intensities persisted after 1 year in the open group: such areas were filled with scar tissue. In the closed group the ultrasound tendon signal returned to normal at 1 year. At 6 months the clinical, radiographic and isokinetic findings did not significantly differ between the open and closed groups. We conclude that defect closure after patellar tendon harvesting does not significantly influence the extensor apparatus.