OBJECTIVE Colonoscopy with air insufflation is known to result in abdominal pain and discomfort. We aimed to determine the effectiveness of water infusion during colonoscopy through a meta-analysis. METHODS Original papers and abstracts published up to October 2011 were searched in MEDLINE, EMBASE, Cochrane Library Database and important meeting abstracts. Clinical appraisal and data extraction were conducted by two reviewers independently. Statistical analysis was performed by meta-analysis using a fixed effects model or a random effects model. RESULTS Seven studies involving 872 patients were included. Meta-analysis showed that the water infusion group had fewer patients requiring abdominal compression or position change [risk ratio (RR) 0.73, 95% CI 0.59, 0.91], a lower mean pain score (RR -1.10, 95% CI -1.26, -0.95), lower maximum pain score (RR -2.34, 95% CI -2.92, -1.76) and fewer patients requiring on-demand sedation (RR 0.45, 95% CI 0.31, 0.66) than the air insufflation group during colonoscopy. There were no significant differences in caecal intubation rate, caecal intubation time, total procedure time and adenoma detection rate. CONCLUSIONS Water infusion significantly decreases patient discomfort and abdominal pain during colonoscopy without affecting operation time and intubation success rate.