Malignant ascites is often refractory to therapy and rapidly deteriorating the nutritional and physical state of the cancer patient. Nevertheless, ascites does not always implicate preterminal state of the cancer process (e.g. ovarian carcinoma). A short review is made of the pathophysiology of ascites in cirrhosis and in malignancy, and different modes of treatment are discussed. The results of medical therapy of malignant ascites (salt and water restriction, diuretics, intraperitoneal cytostatics or radiocolloids) are not convincing. The immunotherapy with OK-432, as worked out by Katano (16-46) has to prove its value. The best and most hopeful results in cases of massive previously resistant ascites, are obtained with a peritoneojugular shunt, improving immediately the nutritional status and life condition, providing excellent palliation. The superiority of the Denver shunt versus the Le Veen shunt has been assessed recently, especially for malignant ascites. Some technical and perioperative details merit more attention, to limit the high risk ratio. Control of the intrathoracic position of the catheter tip, the maintenance of the bloodflow in the jugular vein, the intramuscular tunnelisation of the peritoneal catheter, the discard of 3 or 5 liters ascitic fluid and the substitution of part of it by physiological fluid, perioperative prophylactic antibiotics and heparinisation, flow-rate control in the postoperative period by changing patients position, respiratory exercises, daily flushing, all those measures limit the risk of fibrinolysis (DIC), shunt occlusion, fluid overload and infection. The fear of metastasis by shunt is unfounded, since the survival of the primary tumor is mostly too short (41). The postoperative follow up in an intensive care unit is necessary during 24-72 hours.