Introduction: A better screening method for CP is needed to determine who should undergo more costly and invasive testing. Aims: Develop a CP prediction score to risk stratify patients for treatment or further testing. Methods: Retrospective review of pancreatic disorders database for at-risk patients (alcohol >20 g/day, smoking, RAP, abdominal pain) referred for chronic pancreatic disease evaluation and treatment. Data collection form recorded demographics, symptoms, TIGAR-O risk factors, exocrine dysfunction (pancreatic elastase PE-1 <200 ug/g], secretory dysfunction (peak HC03-<75 meq/L), endocrine dysfunction (abnormal fasting glucose or HA1C) and abdominal imaging findings (Cambridge score on CT, MRI/MRCP). Statistical analysis included univariate (t-test continuous; Fisher’s exact or Chi-squared categorical variables] and multivariate logistic regression analysis [SAS, Cary, NC]. Results: Eighty-eight subjects stratified by pancreas secretory function into 2 groups: normal and abnormal. Groups compared to determine independent predictors of pancreas function. Univariate analysis: The following variables predicted abnormal secretory function: heavy alcohol (p=0.033), nausea (p=0.011), abnormal PE-1 (p=0.009), moderate/marked CT scan (p=.010), or moderate/marked MRI/MRCP imaging (p=0.001). Multivariate analysis: Abnormal secretory function observed more with nausea (OR 11.47; 95% CI 2.13, 61.86), abnormal MRI (OR 9.45; 95% CI 1.25, 71.19), and exocrine insufficiency (OR 6.05; 95% CI 1.18, 31.14). Model formula: −2.206 + 2.246 [MRI] + 2.440 [nausea] + 1.801 [exocrine function]. CP risk score assigns 2 points per variable (total score=0, 2, 4, or 6). The probability of CP (total points; percent predicted CP): low (0 points; 9.5% CP); intermediate (2 points; 50% CP); high (4 points; 86% CP; or 6 points; 100% CP). Overall CP risk score AUC=0.854. Conclusion: A non-invasive prediction score has been derived to stratify at-risk patients for treatment or further diagnostic testing for chronic pancreatitis. Clinical implications: For a low probability score (0), no further pancreas testing. A high probability score (4-6) requires treatment of chronic pancreatitis and its associated complications. An intermediate probability score (2) requires referral for invasive pancreas imaging and physiologic testing.