BACKGROUND: Various sedative drugs have been proposed to control anxiety and agitation in shoulder dislocation, but none of them has been diagnosed as the best sedative and relaxant agent. The study aimed to compare the sedative effectiveness of thiopental versus midazolam in reduction of shoulder dislocation. METHODS: A randomized double-blind controlled trail was performed in 80 patients with shoulder dislocation recruited from the emergency department. Ten patients were excluded and 70 patients were enrolled in the study. Case group received intravenous thiopental 2 mg/kg+2 μg/kg fentanyl while control group received intravenous midazolam 0.1 mg/kg+2 μg/kg fentanyl. Number of times, patients and physician's satisfaction, difficulty of procedure, degree of muscle relaxation, time of sedation and complete recovery, number of patients with apnea episode, O2 saturation, patient's pain score and adverse events were all recorded. RESULTS: Muscular tone had significant difference between the two groups (P-value=0.014) and thiopental was more muscle relaxant than midazolam. Replacement of shoulder dislocation in thiopental group was easier than midazolam group (P-value=0.043). There was no need to use multiple methods of reduction in either group. Before drug infusion the mean±SD VAS scores were 8.37±2.21 in the midazolam group (A) and 8.94±1.78 in the thiopental group (B); mean difference 0.57, 95％CI= –0.38 to 1.52. After completion of the procedure, the mean±SD VAS scores in group (A) and (B) were 3.20±1.30 vs. 3.65±1.30; mean difference –0.45, 95％CI= –1.07 to 0.16. CONCLUSION: Thiopental might be more effective and relaxant than midazolam for reduction of shoulder dislocation.