Our aim in this study was to compare the efficiency of 25G versus 22G needles in diagnosing solid pancreatic lesions by EUS-FNA. We performed a systematic review and meta-analysis. Studies were identified in five databases using an extensive search strategy. Only randomized trials comparing 22G and 25G needles were included. The results were analyzed by fixed and random effects. A total of 504 studies were found in the search, among which 4 randomized studies were selected for inclusion in the analysis. A total of 462 patients were evaluated (233: 25G needle/229: 22G needle). The diagnostic sensitivity was 93% for the 25G needle and 91% for the 22G needle. The specificity of the 25G needle was 87%, and that of the 22G needle was 83%. The positive likelihood ratio was 4.57 for the 25G needle and 4.26 for the 22G needle. The area under the sROC curve for the 25G needle was 0.9705, and it was 0.9795 for the 22G needle, with no statistically significant difference between them (p=0.497). Based on randomized studies, this meta-analysis did not demonstrate a significant difference between the 22G and 25G needles used during EUS-FNA in the diagnosis of solid pancreatic lesions.