This study tests Wolff's law of trabecular bone adaptation by examining if induced changes in joint loading orientation cause corresponding adjustments in trabecular orientation. Two groups of sheep were exercised at a trot, 15 min/day for 34 days on an inclined (7°) or level (0°) treadmills. Incline trotting caused the sheep to extend their tarsal joints by 3-4.5° during peak loading (P<0.01) but has no effect on carpal joint angle (P=0.984). Additionally, tarsal joint angle in the incline group sheep were maintained more extended throughout the day using elevated platform shoes on their forelimbs. A third "sedentary group" group did not run but wore platform shoes throughout the day. As predicted by Wolff's law, trabecular orientation in the distal tibia (tarsal joint) were more obtuse by 2.7 to 4.3° in the incline group compared to the level group; trabecular orientation was not significantly different in the sedentary and level groups. In addition, trabecular orientations in the distal radius (carpal joint) of the sedentary, level and incline groups did not differ between groups, and were aligned almost parallel to the radius long axis, corresponding to the almost straight carpal joint angle at peak loading. Measurements of other trabecular bone parameters revealed additional responses to loading, including significantly higher bone volume fraction (BV/TV), Trabecular number (Tb.N) and trabecular thickness (Tb.Th), lower trabecular spacing (Tb.Sp), and less rod-shaped trabeculae (higher structure model index, SMI) in the exercised than sedentary sheep. Overall, these results demonstrate that trabecular bone dynamically adjusts and realigns itself in very precise relation to changes in peak loading direction, indicating that Wolff's law is not only accurate but also highly sensitive.