Difference in blood and peritoneal glucose (DBPG) is used in clinical practice to support a diagnosis of septic peritonitis in horses. It is inexpensive, easy and rapid to perform. The aim of this study was to evaluate the accuracy of the DBPG to differentiate between septic and non-septic peritonitis in horses. Blood and peritoneal fluids were harvested from suspected animals. Plasma and peritoneal glucose levels, total nucleated cell count, direct microscopic and microbiological examinations of the peritoneal fluid were evaluated. Using DBPG levels, the animals were classified into two groups: difference ≥ 50 mg/dL (positive test) and difference < 50 mg/dL (negative test). Positive microbiological examination and/or presence of bacteria in direct microscopic examination was used as a gold standard to detect septic peritonitis. The accuracy parameters analysed were: sensitivity, specificity, and positive/negative predictive values, for which the results were respectively: 0.23, 0.91, 0.60 and 0.67. Due to poor accuracy, other cut-off margins and peritoneal glucose concentrations were evaluated. The test was considered most accurate when the DBPG was zero with sensitivity, specificity, and positive/negative predictive values of 0.85, 0.82, 0.73, 0.90 respectively. Peritoneal glucose concentrations alone were not a reliable feature to detect peritonitis. DBPG ≥50 mg/dL, widely used for the diagnosis of septic peritonitis, does not have a good accuracy and the DBPG = 0 has a better accuracy for detecting the disease.Copyright © 2020 Elsevier Ltd. All rights reserved.