The mechanical and morphological characteristics of femur-medial collateral ligament-tibia units and the histomorphometry of medial collateral ligament-tibial insertion were examined in female Sprague-Dawley rats with diabetes mellitus (type I, insulin-dependent diabetes). Diabetes was induced with the streptozotocin, a drug with toxic effects on insulin-producing islet cells in the pancreas. The groups studied included rats with untreated streptozotocin-induced diabetes mellitus (Diabetes), insulin-treated streptozotocin-induced diabetes (Diabetes-Insulin), exercise with streptozotocin-induced diabetes (Diabetes-Exercise), exercise with insulin-treated streptozotocin-induced diabetes (Diabetes-Insulin-Exercise), and age-matched sedentary-control rats (Control). Diabetes and Diabetes-Exercise groups had significantly lower body mass and higher blood glucose than Diabetes-Insulin, Diabetes-Insulin-Exercise, and Control groups, indicating that exercise alone did not prevent growth retardation or improve blood glucose control in the streptozotocin-induced diabetes. The strength of Diabetes femur-medial collateral ligament-tibia units was significantly less than Control, but exercise (with or without insulin treatment) maintained the bone-ligament-bone unit strength at a normal level. The load (per unit body mass) of Diabetes-Exercise femur-medial collateral ligament-tibia unit was significantly greater than Control, Diabetes, and Diabetes-Insulin groups. The tensile stiffness (per unit body mass) of the femur-medial collateral ligament-tibia unit for the Diabetes-Exercise group was also significantly greater than Control and Diabetes-Insulin groups. The fibroblast-like cell density in medial collateral ligament at its tibial-insertion site and medial collateral ligament insertion area was significantly less in Diabetes rats. RELEVANCE--:Diabetes mellitus can have deleterious effects on connective tissues, such as bone and ligament. Thus the bone-ligament junction becomes vulnerable. The results of this experiment suggest, however, that exercise, with or without insulin treatment, can have a positive effect on the bone-ligament interface in the rat with type I diabetes mellitus. Furthermore exercise (with insulin treatments) can mitigate growth retardation and many of the adverse effects of streptozotocin-induced diabetes on the structure and mechanical properties of immature bone-ligament junctions.