Objective: The authors evaluated the diagnostic validity of an interview-based panic disorder diagnosis in cardiology chest pain patients with angiographically normal coronary arteries. Method: Patient probands with normal coronary arteries (N=65) were first contacted immediately after their normal angiogram and were given a structured diagnostic interview. On the basis of the results of the interview, probands were grouped as having panic disorder (N=19), panic attacks that did not meet frequency criteria for panic disorder (N=17), or no panic (N=29). At a later time, patient probands were recontacted and given a structured family-history interview that inquired about psychopathology in their first-degree biological relatives (N=544). Results: As predicted, panic disorder was significantly more prevalent among the first-degree relatives of probands with normal coronary arteries diagnosed with panic disorder or panic attacks than among the family members of probands with normal coronary arteries without panic (17.4% versus 15.7% versus 4.0%). Family members of probands with panic attacks were significantly more likely to be diagnosed with major depression than were the family members of probands with no panic; however, differences did not reach significance for family members of the panic disorder proband group. Groups did not differ significantly in familial alcoholism. Conclusions: These data support the construct validity of an interview-based panic disorder diagnosis among patients with chest pain and normal coronary arteries and suggest that these patients could benefit from treatment for panic disorder.