The purpose of this study was to determine the usual circumstances surrounding spinal cord injuries that occur in swimming pools so that appropriate primary prevention programs targeted at high risk persons, activities, and environments could be developed and initiated. A sample of 341 persons enrolled in the National Spinal Cord Injury Statistical Center database since 1973 whose injury was the result of a swimming pool mishap was identi®ed. Medical records were reviewed retrospectively and a survey questionnaire was administered by telephone to 196 persons (57.5%) who were located and agreed to participate. Average age at time of injury was 24 years, 86% of injured persons were men, 95% were white, 70% were never married, 32% had less than a high school education, and only 7% were college graduates. Almost all injuries (87%) occurred in private/residential pools. Most injuries (57%) occurred when diving into less than 4 feet of water, while an additional 38% occurred at water depths between 4 and 8 feet. Depth indicators were not present in 75% of cases. There were no warning signs posted in 87% of cases. There was no lifeguard on duty in 94% of cases. There was self-reported alcohol involvement in 49% of cases, but drugs were involved in only 2% of cases. Almost half of all injuries (46%) occurred during parties. In 44% of cases, the injury occurred during the person's ®rst visit to that particular pool. Ordinary dives accounted for 70% of cases, followed by unusual dives (17%), unintentional pushes (6%), and other circumstances (7%). Almost all injuries (88%) occurred between 1 pm and 1 am with the most frequent time of day being 6 pm, and 82% occurred during June, July or August. Over half (51%) of all injuries occurred on Saturday or Sunday. These results provide important clues to the development of a successful primary prevention program.