The availability of ibuprofen without a prescription requires assessment of its acute toxicity in the general population. We report results from a prospective study of 329 cases of ibuprofen overdose from a statewide cohort that were reported to our poison center between April 1985 and November 1986; 85 patients had ibuprofen serum concentrations measured. Gastrointestinal disturbances and central nervous system depression were the most common clinical findings (42% and 30% of patients, respectively), but the severity was mild; only one patient had severe symptoms that could be attributed to ibuprofen. Ibuprofen serum concentrations correlated poorly with gastrointestinal symptoms (r = -.177), central nervous system findings (r = .176), presence of coingestants (r = .078), and presence of potentially life-threatening symptoms (r = .087). We evaluated the usefulness of a previously published nomogram to predict ibuprofen toxicity; the positive predictive value for severe symptoms was 6% for all patients and 0% for patients ingesting ibuprofen alone. We conclude that the frequency of life-threatening complications from ibuprofen overdose is low, the nomogram is not predictive of toxicity, and routine serum concentration determinations are not useful as an adjunct in the management of overdose cases.