Muscle tone was tested at the shoulders and wrists of 49 randomly selected poststroke patients with the use of resting joint position (SJP and WJP), resistance to passive movement or stiffness (SRM and WRM), and angle of appearance of resistance (SAR and WAR). Subjects were tested while seated with their arm supported in a suspension sling adapted for free movement. Five of the first and immediately repeated measurement pairs showed strong correlations and interrater reliability (SJP, .839; WJP, .900; SRM, .886; WRM, .904; SAR, .884 [p < .05]). The sixth (WAR) showed moderate reliability (.618, p < .05). Resting joint position measurements were most reliable among subjects with higher tone. The joint first measured had a slight order effect on SRM among subjects with higher muscle tone. Its second measurements were slightly increased over the first among those subjects whose shoulders were measured first and slightly reduced when measured immediately after the wrist. Reliable means of clinical evaluation of muscle tone at the shoulder and wrist are available if the influence of level of tone and the mutual influence of muscles tested are prudently considered.