The study's objective was to contrast the prevalence, phenomenology, and medical care utilization for panic disorder from 1980 to 1990. All psychiatric consultations from a university consultation service from the years 1980, 1985, and 1990 were located (N = 2,400). Patients meeting DSM-III-R panic disorder criteria were selected for chart review. Variations in demographics, comorbidity, prior medical evaluations, and referral patterns were analyzed. The prevalence rate for consultations meeting panic disorder criteria has increased (2.5% in 1980, 4.2% in 1985, and 5.1% in 1990; chi 2 = 7.5, P = 0.024). Referring physicians more frequently noted panic attacks in 1990 (5% in 1980, 21% in 1985, and 59% in 1990; chi 2 = 21.2, P = 0.0001). A summary measure of medical care utilization revealed no significant interval change. Panic disorder is being recognized and referred more frequently by medical physicians since the publication of DSM-III and DSM-III-R. However, delay of diagnosis and high medical care utilization remain significant problems.