Objectives: Several morphological risk factors for anterior cruciate ligament (ACL) injury have been identified,1,6,12,13 including the size of the ACL.5,8,12,15 A smaller ACL volume and diameter are associated with a greater risk of injury when comparing ACL-injured subjects to matched controls.5,8,12,15 Although morphological risk factors as a group have been largely characterized as non-modifiable,9,10 ACL surface and cross-sectional areas (CSA) have the potential for modifiability, especially during growth and development.7,11 These ACL area measures have increased and the mechanical properties of the ACL have improved following exercise through periods of growth in animal models.2,7,11,14 In humans, we are only aware of one study of ACL hypertrophy—a small study of elite weightlifters.7The main purpose of this study, therefore, was to determine whether the ACL can hypertrophy in response to mechanical loading by comparing bilateral differences in ACL CSA in athletes who habitually load one leg more than the other in training for their sport. Based on the work of Grzelak et al. in weightlifters,7 as well as animal evidence that the ACL responds to exercise,2,11,14 we hypothesized that these athletes would present with significantly greater ACL CSAs in the landing/drive leg, the knee that is loaded the most in comparison with the contralateral control knee. Demonstrating the potential for the ACL to hypertrophy via mechanical loading would provide a scientific basis for exploring ACL injury prevention strategies aimed at increasing ACL CSA and robustness given that a small ACL volume and diameter are known risk factors for injury. This is particularly important for all children and especially females since they are at a much higher risk for ACL injury, and thereafter the development of knee osteoarthritis. Methods: We recruited 50 figure skaters and springboard divers because they consistently and repeatedly use one leg more than the other, thereby ensuring that one knee was habitually loaded more than the other (Table 1). More specifically, figure skaters always land their jumps on the same leg, while springboard divers always drive the same leg (‘drive’ leg) into the board during their hurdle approach. Sport training for all participants began prior to puberty and continued through and after. Bilateral knee magnetic resonance images (MRIs) were acquired with a Philips Ingenia 3.0-T scanner using a dedicated knee coil. Each knee, resting in slight flexion in the coil, was scanned using three sequences, all in the plane of the ACL: (1) oblique-sagittal (repetition time (TR): 5100 ms; echo time (TE): 30 ms; slice thickness: 2.5 mm; pixel spacing: 0.19 x 0.19 mm); (2) oblique-coronal (TR: 4000 ms; TE: 30 ms; slice thickness: 2.5 mm; pixel spacing: 0.20 x 0.20 mm); (3) oblique-axial (TR: 5100 ms; TE: 30 ms; slice thickness: 2.5 mm; pixel spacing: 0.20 x 0.20 mm). Using the oblique-axial-plane scans, the ACL CSA was measured on the three slices that were closest to 50% of the ligament’s length, and then averaged (Figure 1). Using the oblique-sagittal-plane scans, which were reconstructed to run parallel to the patellar tendon, the anteroposterior diameter of the patellar tendon was measured perpendicular to the tendon’s longitudinal axis at a distance of 2 cm distal to the patella3,4 on the slice displaying the thickest part of the tendon at that height (Figure 2). In addition, isometric and isokinetic knee extensor and knee flexor peak torques were acquired using a dynamometer. Bilateral differences in ACL CSA, PT diameter, and knee muscle strength were evaluated via one-sample t-tests that compared the mean percent difference between limbs to a null hypothesis of a zero mean percent difference. Correlations between bilateral ACL CSA differences, age of training onset and years of training were also examined. Results: Athletes with repeated unilateral lower limb loading had significantly greater ACL CSAs in the dominant knee than the non-dominant knee (Table 1; ACL CSAs: dominant = 42.0 ± 8.8 mm2; non-dominant = 40.8 ± 9.1 mm2; % difference = 4.4 ± 13.8%; t = 2.236; p = 0.030). Also, these athletes had significantly greater AP patellar tendon diameters in the dominant knee than the non-dominant knee (Table 1; patellar tendon diameters: dominant = 4.1 ± 0.6 mm; non-dominant = 3.9 ± 0.5 mm; % difference = 4.5 ± 9.4%, t = 3.322; p = 0.002). The percent bilateral difference in ACL CSA, however, was not associated with training onset (r = 0.087, p = 0.553) or years of training (r = -0.068, p = 0.641). Lastly, isometric knee flexor peak torques were significantly greater in the landing/drive leg than the contralateral knee (Table 2). Peak torques from other contraction types or muscle group did not differ between limbs (Table 2). Conclusions: Athletes who habitually loaded one leg more than the other prior to, during and after puberty exhibited significant unilateral ACL hypertrophy in their landing/drive leg. These results support existing evidence that exercise, including resistance and endurance regimens, during periods of pubertal growth has the potential to increase ACL CSA and, therefore, its strength.2,7,14 This suggests that perhaps the ACL could be ‘trained’ to become larger, more robust, and thus at lower risk of injury given that a smaller ligament is associated with a greater risk of injury. The bilateral difference in patellar tendon morphology supports our assumption that the athletes participating in this study consistently loaded one knee more than the other during their sport training, and that increased loading led to the hypertrophy of two important structures in that knee. Many gaps in knowledge—ACL development during growth and how exercise may alter its morphology and mechanical properties—need to be addressed as injury prevention strategies that involve ‘training’ the ACL are explored.