Background and Purpose - Hemiplegic shoulder pain is not uncommon after stroke. Its origin is still unknown, and although many different methods of treatment are applied, none have yet been proved to be effective. We sought to study the efficacy of 3 injections of intra-articular triamcinolone acetonide on pain and arm function in stroke patients with hemiplegic shoulder pain. Methods - In a multicenter, randomized, placebo-controlled clinical trial, patients with hemiplegic shoulder pain received either 3 intra-articular injections of 40 mg triamcinolone acetonide or 1 mL physiological saline solution (placebo). Primary outcomes were pain measured according to 3 visual analogue scales (score range, 0 to 10), and arm function was measured by means of the Action Research Arm test and the Fugl-Meyer assessment scale; secondary outcomes were passive external rotation of the shoulder and general functioning measured according to Barthel Index and the Rehabilitation Activities Profile. Results - In the triamcinolone group (n=18), the median decrease in pain, 3 weeks after the last injection, was 2.3 (interquartile range, 0.3 to 4.3) versus 0.2 (interquartile range, -0.5 to 2.2) in the placebo group. This result was not statistically significant. The change in the other outcome measures did not differ significantly between the 2 treatment groups. Twenty-five patients reported side effects. Conclusions - In the 37 participants included in this study, triamcinolone injections seemed to decrease hemiplegic shoulder pain and to accelerate recovery, but this effect was not statistically significant. Therefore, on the basis of the results of this study, these injections cannot be recommended for the treatment of patients with hemiplegic shoulder pain.