AIMTo investigate serum mean platelet volume (MPV) levels in acute pancreatitis (AP) patients and assess whether MPV effectively predicts the disease severity of AP.METHODSWe included 117 consecutive patients with AP as the AP group and 34 consecutive patients with colorectal polyps (before endoscopic treatment) as the control group. Complete blood counts, liver function, platelet indices (MPV), coagulation parameters, lactate dehydrogenase (LDH) and C-reactive protein (CRP) were measured on days 1, 2, 3 and 7 after admission. Receiver operating characteristic curves were used to compare the sensitivity and specificity of MPV, white blood cell (WBC), LDH and CRP in predicting AP severity. The Modified Glasgow Prognostic Score (mGPS) and the 2012 revised Atlanta criteria were used to evaluate disease severity in AP.RESULTSMPV levels were significantly lower in the AP group than in the control group on day 1 (P = 0.000), day 2 (P = 0.029) and day 3 (P = 0.001) after admission. In addition, MPV values were lower on day 1 after admission than on day 2 (P = 0.012), day 3 (P = 0.000) and day 7 (P = 0.002) in all AP patients. Based on the mGPS, 78 patients (66.7%) were diagnosed with mild and 39 patients (33.3%) with severe AP. There was no significant difference in mean MPV levels between patients diagnosed with mild and severe AP based on the mGPS (P = 0.424). According to the 2012 revised Atlanta criteria, there were 98 patients (83.8%) without persistent organ failure (OF) [non-severe acute pancreatitis (non-SAP) group] and 19 patients (16.2%) with persistent OF (SAP group). MPV levels were significantly lower in the SAP group than in the non-SAP group on day 1 after admission (P = 0.002). On day 1 after admission using a cut-off value of 6.65 fL, the overall accuracy of MPV for predicting SAP according to the 2012 revised Atlanta criteria (AUC = 0.716) had a sensitivity of 91.8% and a specificity of 47.4% and was superior to the accuracy of the traditional markers WBC (AUC = 0.700) and LDH (AUC = 0.697).CONCLUSIONMPV can be used at no additional cost as a useful, non-invasive biomarker that distinguishes AP with persistent OF from AP without persistent OF on day 1 of hospital admission.