Tonsillectomy is one of the most common surgical procedures, but there is debate whether systemic steroids should be used to reduce pain and post-operative complications. To determine whether peri-operative steroids reduce post-tonsillectomy pain and complications in adults. Systematic review and meta-analysis of randomised controlled trials. We searched MEDLINE (1950-2010), EMBASE (1980-2010), CINAHL (1981-2010), Web of Science, ProQuest, metaRegister, Conference Proceedings Citation Index, the Cochrane Library and reference lists of relevant studies. Two reviewers independently selected trials and extracted data on their quality, characteristics and results. Trials included adults (age >16 years) undergoing elective tonsillectomy where peri-operative steroids were used, and the results were compared with control or placebo. There were seven randomised controlled trials (580 patients) reporting post-operative pain. Meta-analysis demonstrates that dexamethasone in adults reduces the pain level experienced in the first post-tonsillectomy day [standard mean difference (SMD): -0.63, 95% CI: -1.13 to -0.12] with significant heterogeneity (I(2) = 84%, P < 0.00001). Sub-group analysis to explore heterogeneity demonstrated this reduction in pain was mostly with high total dose steroids (total >10 mg over first 24 h post-operatively; SMD: -1.48, 95% CI: -2.17 to -0.79, P < 0.00001), especially when given both intra-operatively and post-operatively. There was no significant effect with low doses (SMD: -0.12, 95% CI: -0.36 to 0.13, P = 0.35). There were three trials (231 patients) that reported post-operative nausea and vomiting, three other trials (270 patients) reporting on bleeding and three trials (401 patients) reporting other complications (infections and odynophagia). There was a significant reduction in post-operative nausea and vomiting (RR: 0.53, 95% CI: 0.36 to 0.80, P = 0.002, I(2) = 26%) and bleeding (RR: 0.45, 95% CI: 0.25 to 0.80, P = 0.007, I(2) = 0%), but the reduction in the other complications did not reach statistical significance (RR: 0.69, 95% CI: 0.48 to 1.01, P = 0.06, I(2) = 0%). Pooling of these complications (post-operative nausea and vomiting, bleeding, infections and odynophagia) shows that in six trials (501 patients), the use of dexamethasone significantly reduced post-operative complications following tonsillectomy in adults (RR: 0.59, 95% CI: 0.49 to 0.71, P < 0.00001, I(2) = 0%), when compared with placebo or control. Dexamethasone reduces pain, post-operative nausea and vomiting, bleeding and overall post-operative complications in adults undergoing tonsillectomy. However, the effect of the dose of dexamethasone on post-operative pain and whether dexamethasone reduces bleeding require further research. © 2011 Blackwell Publishing Ltd.