BACKGROUND: Post-operative shivering is one of the potential complications for any surgical patient. Its incidence varies from 5% to 65%, and many preventive and treatment modalities have been reported. For the effective prevention of post-anesthetic shivering by using intravenous clonidine or meperidine, randomized controlled studies were reviewed. The overall incidence of shivering after clonidine or meperidine administration, and the anti-shivering effect of clonidine and meperidine were evaluated. METHODS: DATA SOURCES: Medline search from 1978 to March 1998. DATA SELECTION: We selected studies that had investigated the preventive anti-shivering effect of intravenous clonidine or meperidine by randomized controlled trials. Ten clinical trials were evaluated. RESULTS: The pooled odd ratio of the patients who received clonidine was 0.32 (95% confidence interval, 0.22~0.47) and it seemed to be effective. But these studies showed little evidence of significant homogeneity (P=0.01). In the subgroup analysis, the pooled odd ratio of group A (early administration or intra-operative infusion group) was 0.47 (95% CI 0.31~0.72) evidenced effectiveness but failed to prove homogeneity (P=0.047). But group B (the late intra-operative administration group) had a pooled odd ratio of 0.10 (95% CI 0.05~0.22) and showed homogeneity (P=0.98). In meperidine trials, the pooled odd ratio was 0.20 (95% CI 0.07~0.55). CONCLUSION: We present quantitative evidence based on a meta-analysis of pooled effect size from randomized trials that clonidine is more beneficial for the prevention of post-anesthetic shivering and more effective than meperidine when it is administrated during later period of surgery.