Objective: To investigate psychiatric and neurological morbidity diagnostic stability and indicators of prognosis in patients previously identified as having London medically unexplained motor symptoms. Design: Follow up study. Setting: National Hospital for Neurology and Neurosurgery; London-a secondary and tertiary referral hospital for neurological disorders. Subjects: 73 patients with medically unexplained motel symptoms admitted consecutively in 1989-91. 35 (48%) patients had absence of motor function (for example, hemiplegia) and 35 (52%,) had abnormal motor activity (for example, tremour dystonia, or ataxia). Main outcome measures: Neurological clinical diagnosis at face to face reassessment by a neurologist and a psychiatric diagnosis after a standardised assessment interview-the schedule for affective disorders and schizophrenia-conducted by a psychiatrist. Results: Good follow up data were available for 64 subjects (88%). Only three subjects had new organic neurological disorders at follow up that fully or partly explained their previous symptoms. 44/59 (75%) subjects had had psychiatric disorders; in 33 (75%) patients, the psychiatric diagnosis coincided with their unexplained motor symptoms. 31/59 (45%) patients had a personality disorder: Three subjects had developed new psychiatric illnesses at follow up, but in only one did the diagnosis account for the previous motor symptoms. Resolution of physical symptoms was associated with shaft length of symptoms, comorbid psychiatric disorder, and a cl-range in marital status during follow up. Conclusions: Unlike Slater's study of 1965, a low incidence of physical or psychiatric diagnoses which explained these patients' symptoms or disability was found. However, a high level of psychiatric comorbidity existed.