BACKGROUND: Endoscopic Ultrasound (EUS) is considered the most sensitive method for diagnosing chronic pancreatitis; however poor reproducibility has been reported by some. On the other hand, histogram analysis of echogenicity is an objective assessment of the pancreatic parenchyma and theoretically could be more reproducible without involving the subjective evaluation of the observer. Also some of the established criteria can be included in this analysis (such as heterogeneity, lobularity and stranding) MATERIAL AND METHODS: Two study groups were formed: Group 1: asymptomatic patients with no known pancreatic disease and “normal” pancreas by EUS, Group 2: patients with chronic pancreatitis established by other methods (mainly ERCP). EUS was performed using a Pentax FG32UA linear array echoendoscope coupled to a Hitachi EUB 515 console. Histogram analysis was performed in a 20mm2 area in the body of the pancreas without including the duct. This was done upon entering at the start of the study and then on the way out, to establish reproducibility. Chronic pancreatitis was diagnosed when more than three established criteria were found (excluding duct criteria). The Histogram variables studied included mean and standard deviation values. Reproducibility for each patient was assessed by Pearson's Correlation Coefficient (histogram) and Kappa (Standard criteria). Values between groups were compared with Student's T test. A ROC curve was constructed to determine a cut-off point for mean histogram values at which this method could provide the most sensitivity and specificity. RESULTS:We have included 130 patients, mean age 49 (19-86), other (table): Reproducibiliy for histogram:.90, for standard criteria:.81 , both p <0.05. Using 35 as cut-off value: sensitivity: 86%, specificity: 75%. Younger and leaner patients tended to have less echogenic parenchyma than older or obese individuals. CONCLUSIONS: Histogram analysis is similarly sensitive and specific as standard criteria but may be more reproducible.