To increase in vivo tendon force and gliding alter flexor tendon repair, a variety of modifications to the methods by which protective passive motion is administered have been advocated. To determine the relationship between the prime variables, wrist and digital position, muscle activation, and in vivo tendon force, a clinically relevant canine model was developed. Force was measured in that flexor tendon during several joint manipulation paradigms: single-finger flexion-extension with the wrist flexed (group 1F), single-finger flexion-extension with the wrist extended (group 1E), four- finger flexion-extension with the wrist flexed (group 4F), four-finger flexion-extension with the wrist extended (group 4E), and synergistic wrist and finger motion where wrist extension and finger flexion were performed simultaneously, followed by wrist flexion and finger extension (group SYN). In addition, tendon force was measured during electric stimulation of the proximal flexor muscle mass. Passive tendon force with the wrist extended (groups 1E and 4E) was two to three times greater than that measured with the wrist flexed, independent of the number of digits moved. With the wrist extended, peak tendon force reached 1,977 g ± 194 g during single digit manipulation (group 1E), compared to only 853 g ± 104 g with the wrist flexed during the same maneuver (group 1F). Statistical comparison between means revealed that groups 1E and 4E were significantly different from groups 1F, 4F, and SYN (p < .005). There were no significant differences between groups 1E and 4E or between groups 1F, 4F, and SYN (p > .200). Active muscle force elicited by electrical stimulation and passive force varied dramatically as the wrist was flexed from full extension 3460 g ± 766 g to full flexion 427 g ± 239 g (p < .001). Simultaneously, passive tension decreased from 940 g ± 143 g with wrist extended to 76 g ± 37 g with the wrist flexed. These data indicate that wrist position has the greatest effect on flexor tendon force during motions that are commonly used to rehabilitate flexor tendon repairs. Thus, if force is to be controlled during passive motion, wrist-joint angle will have the dominant effect, while the number of digits manipulated will have much less of an effect. If the clinical goal is to minimize tendon force, rehabilitation could be carried out with the wrist flexed, whereas if the goal is to increase tendon force, rehabilitation could include exercise programs that use a greater degree of wrist extension.