Background: Poisoning continues to be an important public health problem in the US. In 1995, 2 million human poison exposures were reported to all poison centers in the US. Hospital emergency department data may be used to examine the most critical nonfatal poisoning exposures. Methods: Data from the 1993-1996 National Hospital Ambulatory Medical Care Survey, which is a national probability sample survey of visits to emergency departments of nonFederal, short-stay, and general hospitals, were examined to describe poisoning-related emergency department visits in the US. Results: During 1993-1996, the average annual number of emergency department visits was 93 million, of which 37 million were injury related and 1 million were poisoning related. Children under 5 years of age had a significantly higher average annual rate of poisoning-related visits (84 visits per 10,000 persons) than persons 5-19 years of age and persons 35 years of age and over. “Poisoning by other and unspecified drugs and medicinal substances” was the leading diagnosis and was recorded at 21% of all poisoning-related visits. Poisoning-related visits were more often recorded as urgent (75%) and were more likely to result in hospital admission (22%) compared to illness visits (45% and 17%, respectively) and nonpoisoning-related injury visits (47% and 6%, respectively). Conclusions: Poisoning-related injury visits comprise a small (1%), but important component of the health care provided in emergency departments. An examination of different definitions of poisoning revealed that for emergency department data, it is most appropriate to use the poisoning E-codes from the injury data framework developed by the injury control community. Data from emergency departments are needed to monitor any changing patterns of nonfatal poisonings and to provide guidance for effective poison prevention programs.