The intention of this controlled family study was to evaluate reasons for comorbidity of schizophrenia and panic disorder. Observed rates of psychiatric disorders in first-degree relatives of patients and of controls were compared with rates predicted by possible hypotheses explaining comorbidity. The sample consisted of 59 patients with schizophrenia (including seven with schizophreniform disorder), 54 patients with panic disorder (with or without agoraphobia), 29 comorbid patients with lifetime diagnoses of panic disorder and schizophrenia (or schizophreniform disorder, 2 patients) and 109 controls, and their 1068 first-degree relatives. Information from clinical performance, clinical and structured interviews, and from family history was joined to establish DSM-III-R diagnoses in patients and relatives. As expected, schizophrenia and panic disorder were found to be familial. The hypothesis, that the familial load for primary panic disorders distinguished schizophrenics (4.3%) and controls (0.9%), could be verified (P < 0.01); the familial aggregation for primary panic disorders did not distinguish schizophrenics and subjects with panic disorder. However, the risk for schizophrenia was not enhanced in relatives of patients with panic disorder (0%) in comparison to controls (0.3%, P > 0.05). The observed familial aggregation pattern of psychiatric disorders in relatives of schizophrenics, panic patients, comorbid patients, and controls refers to an etiological relation of schizophrenia and panic disorder, or at least to a relationship of subgroups of these disorders.