Malignancy-related ascites gives rise to troublesome symptoms and carries a poor prognosis. Abdominal paracentesis is a widely used and effective procedure for symptom relief, but practice with regard to the procedure itself may vary considerably. Evidence obtained in the context of liver disease probably influences practice in malignancy, although the pathophysiology involved is different. Anecdotal evidence suggested a difference in practice between the Exeter and District Hospice and the adjoining hospital, and this was confirmed by a review of case notes. Patients undergoing paracentesis in the hospital were more likely to have prior ultrasound assessment and to be given intravenous fluids, and had longer drainage times and longer inpatient stays. It seemed that some practices were placing unnecessary burdens on patients whose life expectancy was short. A set of clinical guidelines for the procedure was drawn up, based on the limited evidence available and the practice within the hospice. These guidelines emphasise performing ultrasound investigations only in cases of diagnostic uncertainty, allowing up to 5 L of fluid to drain without clamping, leaving drains in for no more than 6 h and giving intravenous fluids only when specifically indicated. These guidelines were introduced on the oncology ward of the adjoining hospital. The impact on practice was assessed by means of a retrospective case note review of all procedures carried out on the ward in the 6 months before and after the guidelines were introduced. The introduction of guidelines resulted in significant reductions in prior ultrasound assessment, mean length of time drains were left in and mean length of inpatient stay for planned procedures. There were no cases of symptomatic hypotension in the postguidelines group.