1 Data from the Boston Collaborative Drug Surveillance Program were reviewed to determine the proportion of patients receiving intravenous fluids in the participating hospitals. 2 Wide differences between hospitals in the United States and four other countries were observed: i.v. fluids being given to 54% of patients in one American hospital and only 7% of patients in an Israeli one. A two‐fold difference in the frequency of i.v. fluid use in two otherwise comparable Scottish teaching hospitals was observed. This difference was not due to observed patient characteristics, did not arise from selection bias or observational differences between the two hospitals and was unlikely to have arisen by chance. 3 It is concluded that the findings were due to different policies on the part of the attending physicians. Although the study could not be used to evaluate the beneficial effects of the administered fluids, adverse effects were common (15% of recipients) and in some instances potentially serious. Adverse effects were reported more frequently after infusion of 5% dextrose (13% of recipients) than after isotonic saline (7%) perhaps because of the low pH of the former solution.