Alarms in the operating room remain a major source of annoyance and confusion. Nearly all alarms result from a transgression of certain alarm limits. We surveyed manufacturers at a major meeting of anesthesiologists for their default alarm limits. We also conducted a mail survey of anesthesiologists in the United States, Finland, and The Netherlands, to learn their chosen alarm limits, reasons for turning off the alarms, and estimates of the number of false alarms. The surveys of medical monitoring manufacturers demonstrated a wide variety of default alarm settings. Anesthesiologists' chosen alarm limits tended to parallel the manufacturers' defaults. Fewer than 30% of anesthesiologists stated that they did not turn off the alarms. The leading reason for turning off alarms was the large number of false alarms. Estimates of the number of false alarms varied from 30% to 76%. The desired role of alarms in anesthetic practice and some suggestions for the general improvement of alarms are discussed.