Does neuromuscular training improve knee function more than traditional strength training following anterior cruciate ligament (ACL) reconstruction? Randomised controlled trial. Two outpatient rehabilitation clinics in Norway. 74 adults aged between 16 and 40 years (mean: 28 years), scheduled for arthroscopic reconstruction of the ACL using an autogenous bone-patellar tendon-bone graft. Exclusion criteria included age of ACL injury more than 3 years, meniscal damage requiring repair, or previous injury or surgery to either knee. Participants were allocated to one of two rehabilitation programs using concealed allocation. A 6-month rehabilitation program was commenced during the second week after surgery following a home program to restore knee range of movement and reduce swelling. Exercises were supervised by physiotherapists during clinic visits twice per week. Knee braces were not used at any time. The neuromuscular training group performed balance exercises, plyometric exercises, agility drills, and sportspecific exercises. The strength training group performed mainly strengthening exercises of the lower extremity (quadriceps femoris, hamstring, gluteus medius, and gastrocnemius muscles) based on the American College of Sports Medicine guidelines. Participants were assessed preoperatively and at 3 and 6 months postoperatively. The primary outcome measure was the Cincinnati Knee Score (CKS) which has a scale of 0 to 100 (100 = normal knee). Secondary outcome measures included pain intensity (100 mm VAS where 0 = no pain, 100 = worst imaginable pain), global knee function (100 mm VAS where 0 = worst possible knee function, 100 = preinjury knee function), isokinetic muscle strength, the 36-Item Short-Form Health Survey, hop tests, proprioception, and balance tests. The outcomes assessor was unaware of treatment allocation. 89% of participants underwent the 6-month assessment. At this time the CKS score was higher in the neuromuscular training group compared with the strength training group by 7 points (95% CI 2 to 13). Similarly, global knee function VAS scores at 6 months were higher in the neuromuscular training group by 13 mm (95% CI 2 to 24). The groups did not differ significantly on the pain VAS or other outcomes. Adherence to the rehabilitation program was higher in the strength training group (91% adherent) compared with the neuromuscular program (71%). Neuromuscular rehabilitation results in superior knee function at 6 months after ACL reconstruction compared with standard strength training.