Background and study aims Endoscopic necrosectomy is limited by the proximity of necrosis to the gastrointestinal tract. Percutaneous endoscopic necrosectomy (PEN) is a minimally invasive endoscopic method of percutaneous debridement. Studies regarding its efficacy and safety are lacking. The purpose of this study was to assess the efficacy and safety of PEN in necrotizing pancreatitis. Methods Pubmed, Ovid, Cochrane, Scopus and Web of Science Database were searched from inception through February 2021. Dual extraction and quality assessment of studies using Cochrane risk of bias tool were performed independently by two authors. The primary outcome was defined as clinical success of PEN. Secondary outcomes included periprocedural morbidity, mortality, and long-term morbidity and mortality. Results Sixteen observational studies including 282 subjects were analyzed. The average reported age of the participants was 50.3 years. Patients with reported gender included 39 % females and 61 % males. The success rate as defined by complete resolution of necrosis and removal of drainage catheters/stents was 82 % (95 % confidence interval 77–87). The mean size of pancreatic necrosis was 14.86 cm (5–54 cm). The periprocedural morbidity rate was 10 %, while there was no reported periprocedural mortality. The long-term morbidity rate was reported as 23 % and mortality at follow-up was 16 %. Conclusions PEN is a novel method of endoscopic management of pancreatic necrosis. Based on our meta-analysis of retrospective studies, it represents a safe treatment modality with high rates of clinical success and low rates of perioperative morbidity and mortality. This study supports the use of PEN when conventional endoscopic therapy is not feasible.