OBJECTIVES: This systematic review assessed whether Tuina (therapeutic massage) is more effective and safer than no treatment or routine medical treatment for irritable bowel syndrome (IBS). METHODS: Eleven databases were searched for randomized controlled trials of IBS diagnosed based on Manning or Rome criteria. Tuina with or without routine treatments (RTs) was tested against RTs. The Cochrane risk of bias was evaluated for each trial. RevMan 5.3 was used to conduct a meta-analysis. RESULTS: A total of 8 trials (5 IBS-diarrhea and 3 IBS-constipation) with 545 participants using 8 different manipulations were included. All trials were published in Chinese. For overall symptom improving rate (> 30 % improvement in overall symptom scores), it had not been shown that Tuina was significantly better than RTs (RR 1.23, 95 % CI 0.94-1.60, 197 participants, 3 studies, I = 65 %) for IBS-diarrhea, and Tuina combined with RTs showed more benefit than RTs alone (RR 1.29, 95 % CI 1.08-1.54, 115 participants, 3 studies) for IBS-diarrhea. All trials did not report adverse effect in relation to Tuina. Risk of bias was generally unclear across all domains. CONCLUSIONS: Tuina combined with RTs may be superior to RTs for improving overall symptom of IBS-diarrhea. Due to the existing methodological issues and the heterogeneity of Tuina manipulation, current findings need to be confirmed in large scale, multicenter, and robust randomized trials (especially on outcome assessing blinding and allocation concealment).