To examine the influence of hospital paging systems on residency training, nursing services, and patient care, we asked medical interns (first-year residents) in three teaching hospitals to keep logs of pages they received during a three-day period. Thirty-one logs from 26 interns were completed; a total of 1206 pages were recorded on 91 days (1095 hours). Interns were paged an average of once an hour; on 24 occasions, interns were paged five or more times an hour. The majority of pages (65 percent) occurred when interns were engaged in patient care. Only 34 percent of the pages were judged both to require a response within one hour and to result in a change in patient care. Twenty-four percent were clinically indicated and required a response within one hour but did not result in a change in patient care. Sixteen percent of pages resulted in a change in patient care or were clinically indicated but could have been postponed for more than an hour. An additional 26 percent of pages neither resulted in a change in clinical management nor were clinically indicated. Reducing the number of unnecessary pages and postponing nonurgent ones could result in as much as a 42 percent decrease in disruptions of patient care and more rest for interns.