Malignant ascites is relatively common in patients with certain types of end-stage cancer. Traditional treatments based on fluid and salt restriction and diuretic therapy often are not able to contain neoplastic ascites. These patients consequently undergo repeated abdominal paracentesis, with further plasma protein loss and risk of injury to abdominal viscera. The aim of this study was to evaluate our experience with Denver peritoneovenous shunt and the outcome of patients with malignant ascites and suggest some modifications to improve device patency. From February 1997 to December 1999, 44 Denver peritoneovenous shunts were placed in 42 patients, 17 women and 25 men, aged between 38 and 77 years (mean, 62.3), affected with malignant ascites due to advanced abdominal cancer. At the time of admission, 72% of patients had pain, 88% dysphagia, 66% nausea and/or vomiting, and 83% dyspnea. Eleven patients underwent local anesthesia with lidocaine and 33 general anesthesia with rapidly metabolized drugs. In 27 cases we used the peritoneal-internal jugular right vein surgical approach and in 3 cases the peritoneal-femoral vein surgical access, joining the saphena vein to the cross. In 10 cases, a radiological positioning of the Denver peritoneovenous shunt was effected by a trans-subclavian access. Relief of ascites symptoms was obtained in 87.5% of cases, with reduction of dyspnea, an increased appetite and improved ambulation. Denver peritoneovenous shunt is a good device to relieve malignant ascites, thereby reducing the risk of complications and the number of hospital admissions due to repeated paracentesis and consequently improving the quality of life. A careful patient selection, an accurate follow-up and some device modifications could improve the shunt performance, allowing a wider application of the Denver peritoneovenous shunt.