A prospective study identified 45 patients with malignancy-related ascites among 448 ascites patients (10% of the total). Patients were categorized into five subgroups based on the pathophysiology of ascites formation. Each subgroup had a distinctive ascitic fluid analysis. Patients with peritoneal carcinomatosis but without massive liver metastases (53.3% of the patients with malignancy-related ascites) had a uniformly positive ascitic fluid cytology, high ascitic fluid protein concentration and low serum-ascites albumin gradient. Patients with massive liver metastases and no other cause for ascites formation (13.3% of the series) had a negative cytology, low ascitic fluid protein concentration, high serum-ascites albumin gradient and markedly elevated serum alkaline phosphatase. Those with peritoneal carcinomatosis and massive liver metastases (13.3% of the series) had a nearly uniformly positive ascitic fluid cytology, variable protein concentration, high serum-ascites albumin gradient and markedly elevated serum alkaline phosphatase. Chylous ascites (6.7%) was characterized by a milky appearance, negative cytology and an elevated ascitic fluid triglyceride concentration. Patients with hepatocellular carcinoma superimposed on cirrhosis (13.3%) had negative ascitic fluid cytology, low ascitic fluid protein concentration, high serum-ascites albumin gradient and elevated serum and ascitic fluid alpha-fetoprotein concentration. Two-thirds of patients with malignancy-related ascites had peritoneal carcinomatosis; 96.7% of patients with peritoneal carcinomatosis had positive ascitic fluid cytology. Ascitic fluid analysis is helpful in identifying and distinguishing the subgroups of malignancy-related ascites.