To evaluate the effectiveness of a neuromuscular warm-up program in preventing acute knee injury in adolescent female football (soccer) players. Cluster randomized (by team) controlled trial, stratified by geographical district. Sample size was calculated (n = 8118) with 80% power to show a reduction of 50% in an estimated 1.15% annual incidence of anterior cruciate ligament (ACL) injury at P ≤ 0.05. Clubs in 8 regional districts of the Swedish Football Association, during the 2009 season (April through October). Female under-14 to under-18 football clubs (ages 12-17 years) were recruited. Reasons for the exclusion of clubs were lack of response, <2 training sessions per week, and the current use of an injury prevention program. The clubs were randomized to a neuromuscular warm-up intervention (Knäkontroll, SISU Idrottsböcker, Sweden, 2005) or to a control group, who were instructed to continue with their usual training and playing practices. The neuromuscular training program included 6 exercises that focused on knee control and core stability (1- and 2-legged knee squats, a pelvic lift, the bench, the lunge, and jump/landing). The exercises were to be done twice per week and were to take about 15 minutes, after a brief running warm-up. They progressed through 4 levels of difficulty. The team coaches supervised the program after instruction from study therapists. The primary outcome was the rate of ACL injuries. Diagnosis was confirmed, as appropriate, by a physician and by magnetic resonance imaging. Secondary outcomes were the rates of serious knee injury and any acute knee injury, defined as those with sudden onset during play that led to a player being unable to participate in training or competition. Severe injuries were those that caused absences of >4 weeks. Two study therapists evaluated the injuries. The coaches recorded data, including when the intervention was performed, any injuries, individual playing times, and periods of absence. Assessment of the primary outcome was done by physicians blinded to group assignment. During 278 298 hours of play, 96 knee injuries occurred in 92 players (intervention group 48, control group 44). The rate did not differ between groups. Of the 21 ACL injuries, 7 occurred in the intervention group and 14 in the control group, giving a rate ratio (RR) of 0.36 (95% confidence interval [CI] 0.15-0.85; P = 0.02). Severe injuries (intervention group 26, control group 31) did not differ between groups. They included 22 collateral or capsular sprains, 21 ACL injuries, 7 patella dislocations or subluxations, 6 meniscal or chondral lesions, and 1 tibial plateau fracture. Compliant players (those who performed ≥ 1 exercise session per week; 1303 players) had a lower rate of ACL injury (RR, 0.17; 95% CI, 0.05-0.57), of severe knee injury (RR, 0.18; 95% CI, 0.07-0.45), and of any acute knee injury (RR, 0.53; 95% CI, 0.30-0.94) than the control group. A short weekly neuromuscular exercise program reduced the rate of ACL injuries among adolescent female football (soccer) players. Those who were compliant with the intervention had fewer severe knee injuries and fewer injuries overall.