It has been suggested that osteonal remodeling is triggered by bone microdamage. The validity of this theory rests on the assumption that loading within the physiological range will produce substantial microdamage with relatively few load cycles. The object of the first experiment was to determine threshold values required to consistently produce fatigue microdamage in vivo. The left forelimb of five groups of dogs, characterized by different strain levels and different numbers of load cycles, were loaded in three point bending. The number of microscopic fields which contained some microdamage was calculated as a percentage of the total number of fields. This experiment indicated that loads producing strains as low as 1500 microstrain on the radius and 1400 microstrain on the ulna for 10,000 cycles will produce significant bone microdamage. A second experiment was performed to verify this threshold and to determine whether microcracks are associated with the initiation of bone remodeling. Procedures in this experiment were the same as those in the first, except that all dogs were loaded in such a way as to produce strains on the radius of 1500 microstrain for 10,000 cycles, and the dogs were sacrificed 1-4 days after loading. The loaded limb demonstrated significantly more microdamage than the control limb (p = 0.03). Moreover, we observed 44 times as many microcracks in association with resorption spaces as expected by chance alone. These data support the hypothesis that fatigue microdamage is a significant factor in the initiation of intracortical bone remodeling.