To identify etiology, clinical findings, diagnostic results, treatment, and short-and long-term survival and to report factors associated with nonsurvival and survival in horses with peritonitis. Study design: Retrospective study. Animals: Horses (n = 72). Methods: Medical records at William R. Pritchard Veterinary Medical Teaching Hospital from 2007-2017 were reviewed for horses diagnosed with peritonitis. The essential inclusion criterion was a peritoneal nucleated cell count of ≥25 000 cells/μL. Gastrointestinal rupture and cases in which peritonitis occurred after abdominal surgery or castration were excluded. Information retrieved from medical records included signalment, history, clinicopathological and peritoneal fluid variables, diagnostic imaging findings, inciting cause, treatment, and short-and long-term survival. Data were analyzed by using Fisher's exact test, Wilcoxon rank sum test, and χ 2 test (P < .05). Results: Colic was the most common presenting complaint (34/72 [48%]). A definitive diagnosis could be made in 44 (44/72 [61%]) cases. The most common cause of peritonitis was infectious agents (31/72), followed by trauma (8/72), gastrointestinal thickening (3/72), and eosinophilic peritonitis (2/72). Idiopathic peritonitis was identified in 28 (28/72 [39%]) cases. Sixty (83%) horses survived to hospital discharge. Long-term follow-up was available for 49 horses, with 43 (88%) horses alive 1 year after discharge. Significant differences between nonsurvivors and survivors were history of colic, positive peritoneal fluid culture, and several hematological/peritoneal fluid variables. Conclusion: Peritoneal fluid analysis was essential for a definitive diagnosis of peritonitis, and certain variables were useful for predicting outcome.