Introduction: 1. A major long-term determinant of anterior cruciate ligament (ACL) graft strength is the biological remodeling process to which the ACL graft collagen matrix is subjected immediately after surgical implantation. The collagen matrix provides the tensile strength of ligaments. Previous quantitative ultrastructural studies have shown that both human and goat anterior cruciate ligament patellar tendon (ACL-PT) autografts contain mainly small diameter collagen fibrils (< 100 nm) at 3–20 years post-surgery irrespective of the tissue of origin of the graft. Similar observations have recently been reported for human ACL allografts over an 8 year period.1 2. Parry and Craig (1988) suggested there was a strong correlation between large collagen fibrils and the ultimate tensile strength and modulus of ligaments/tendons. This part of the study was designed to test the hypothesis that there was a positive correlation between large collagen fibrils and the tensile modulus and the area-normalized ultimate load of patella tendon preparations subjected to loading and unloading conditions. Methods: Two studies are presented in review. 1. This part of the study examined goat ACL-PT autografts up to 3 years post-surgery for collagen fibril size, collagen type and hydroxypyridinium (HP) cross-link density. Twenty-two mature female goats received an ACL-PT autograft to the right knee and were tested at 6 weeks (n = 5), 12 weeks (n = 4), 24 weeks (n = 5), 1 year (n = 5) and 3 years (n = 3). Two in each group were assigned for HP analyses and the rest for collagen fibril quantitation. The fibrils of five left control ACL’s and two PT’s were also quantitated. Two normal animals served as controls for HP analyses. 2. Experimental design. Nineteen Flemish giant rabbits (average weight 5.2 ± 0.6 kg) were used in this second study. Three experimental conditions were tested in which the right limbs were used as the experimental side and the left as the control side. 1. A central window of 40% of the width of the patella tendon was removed and the animals (n = 4) sacrificed immediately (T = 0), 2. Same surgery as in 1. but sacrificed at 7 weeks (n = 6), 3. Same surgery as in 2. but a stress shielding device implanted to unload the tendon (n = 6) and sacrificed at 7 weeks. Groups 2. and 3. were allowed 1 week cage activity and were then exercised on a treadmill for 6 weeks and then sacrificed at 7 weeks post-surgery (T = 7). The patella tendons were tested mechanically for area-normalized ultimate load, tensile stiffness and corresponding modulus from the linear range of response were determined for each tendon. Tranverse sections of the same tendons were prepared for computerized quantitative collagen fibril analyses after tensile testing. The mean of the area weighted mean collagen fibril diameters was determined from four tissue blocks obtained from the central region of each control and experimental patella tendon mechanically tested. These mean fibril diameters were then used in regression correlation analyses vs modulus, stiffness and ultimate load at T = 0 and T = 7 for both the control and experimental tendons. Results: 1. The collagen fibril profile for each time group was compared to the control ACL and PT with Kolomogorov–Smirnov analyses. Differences were found between the PT collagen profile with the 12, 24 and 52 week groups (P = 0.005, 0.07 and 0.1, respectively). All the grafts except the 6 week group contained mainly small fibrils ( 100 nm with Young’s modulus of the grafts which was tested in a separate study. The HP density was low in the 6, 12 and 24 week groups, but increased in the 1 and 3 year groups. Significant correlation (r = 0.75, P = 0.01) was found between the HP density and Young’s modulus.1 2. Strong positive correlations were found between the area weighted mean fibril diameter and the tensile modulus (r = 0.74, P < 0.05) and the normalized ultimate load (r = 0.72, P < 0.05) for the whole rabbit control tendons and the experimental tendons. Discussion: 1. This first study shows that the poor long term strength of goat ACL-PT autografts is probably due to the lack of large fibrils and the low cross-link density determined in the predominantly small collagen fibril populations in these grafts even at 3 years post-surgery.2 2. This is the first study to successfully correlate area weighted mean collagen fibril diameters with biomechanical parameters such as ultimate load, tensile stiffness and modulus of the patella tendon.3 References 1Shino K , Oakes BW et al. (1995) Am J Sports Med 23, 203–9. 2Ng GYF , Oakes BW , Deacon OW , McLean ID , Eyre D (1996) J Orthop Res 14, 851–6. 3Oakes BW , Singleton C , Haut R (1998) TORS, New Orleans, USA 23, 24–4.