Continuous regional arterial infusion (CRAI) is a drug delivery system, which dramatically increases the drug concentration in the pancreas. Previous clinical and basic studies have demonstrated the possible therapeutic efficacy of CRAI for severe acute pancreatitis (SAP). This meta-analysis of all published randomized controlled trials (RCTs) was conducted to assess the efficacy and safety of CRAI for the treatment of SAP. Up to August 10, 2014, RCTs comparing CRAI with intravenous infusion for SAP in PubMed, Embase, EBSCO, MEDLINE, Science Citation Index Expanded, Cochrane Library, China Academic Journals Full-Text Database, Chinese Biomedical Literature Database, and Chinese Scientific Journals Database were selected by two independent reviewers. The relative risk (RR) and their 95% confidence intervals (CI) for duration of elevated serum amylase and urine amylase, duration of abdominal pain, infection rate, incidence of complication, overall mortality, curative rate, hospital stay and details of subgroup analysis were extracted. Meta-analyses were made using the software Review Manager (RevMan version 5.10). Six RCTs with 390 patients meeting the inclusion criteria were included in the final analysis. Compared with intravenous infusion route, CRAI significantly shortened the duration of elevated urine amylase (MD=-2.40, 95% CI=-3.20, -1.60; P<0.00001) and the duration of abdominal pain (MD=-1.46, 95% CI=-1.94, -0.98; P<0.00001), decreased the incidence of complication (RR=0.35, 95% CI=0.15, 0.81; P=0.01) and overall mortality (RR=0.25, 95% CI=0.08, 0.78; P=0.02), shortened the duration of hospital stay (MD=-10.36, 95% CI=-17.05, -3.68; P=0.002), and increased the curative rate (RR=1.66, 95% CI=1.13, 2.46; P=0.01). No mortality and catheter-related infections due to CRAI administration was reported in these studies. Subgroup analysis showed that the combination of drug administration via CRAI did not significantly improve the outcomes. CRAI is effective for the treatment of SAP, and the combination of drug administration via CRAI did not have a significant effect on the improvement of the outcomes.