Neurotic syndromes are defined by characteristic patterns of symptoms, but the validity of the distinction between one syndrome and another depends on associations between the syndromes and clinical history, or treatment response factors that are independent of the defining phenomena. In both a group of twin volunteers and a group of patients with panic disorder/agoraphobia, the lifetime experience of more than one diagnosis of a neurotic syndrome was common but there was no evidence of patterns of co-occurrence of diagnoses being associated with particular syndromes. Receiving a diagnosis was associated with abnormal scores on measures of neuroticism and locus of control, the extent of the abnormality increasing with the number of different diagnoses satisfied. It is argued that the concept of a general neurotic syndrome depends in part on the presence of such predisposing personality factors, and that reduction in this predisposition to neurosis should be the focus of treatment.