An observational, historical cohort evaluation was performed to examine the hypothesis that terfenadine (Seldane) exposure increases the risk of developing life-threatening ventricular arrhythmias. The study population consisted of Medicaid recipients from 4 states that were included in the Computerized On-Line Medical Pharmaceutical Analysis and Surveillance System (COMPASS). The drug exposure period was defined prospectively as 30 days in all treatment cohorts. The primary end point was the development of life-threatening ventricular arrhythmias (ventricular tachycardia, fibrillation and flutter, and cardiac arrest and sudden death). The comparison cohorts included terfenadine (n = 181,672), over-the-counter antihistamines (n = 150,689), ibuprofen (n = 181,672) and clemastine (Tavist; n = 83,156). Over the exposure period, a total of 317 life-threatening ventricular arrhythmic events occurred, 244 of which were cardiac arrests. The incidence of total life-threatening ventricular arrhythmic events and cardiac arrests were more frequent in patients receiving over-the-counter antihistamines (relative risk 0.36) than in those receiving terfenadine, a finding that was consistent across all subgroups. There was no increased risk of life-threatening ventricular arrhythmias in the terfenadine cohort as compared with the ibuprofen cohort (relative risk 0.62), and in some analyses, the ibuprofen cohort had a significantly higher arrhythmic event rate. In all comparisons with the clemastine cohort, the terfenadine cohort had a statistically indistinguishable relative risk (1.08). Age, race, sex and cardiovascular risk were all considered in the adjusted relative-risk analyses. No baseline historical characteristic or imbalance of baseline medications explained the differences between groups.(ABSTRACT TRUNCATED AT 250 WORDS)