We describe a personal series of 60 cases of parkinsonism with onset under the age of 40 years. Known causes for early onset of secondary parkinsonism, such as Wilson's disease or encephalitis, were excluded in every case. Two groups were identified: those with onset after the age of 21 in whom no hereditary factors could be ascertained (56 cases), and those with onset before 21 years all of whom had familial parkinsonism. In neither group have we found any association with prematurely grey hair, hypertension, diabetes, pernicious anaemia, or thyroid disorder. Among their families, we have not found any association with diabetes, pernicious anaemia, or thyroid disorder. We propose that cases of apparent idiopathic Parkinson's disease beginning between age 21-40 years should be called "young onset Parkinson's disease." Twenty percent of such patients in our series had at least one first- or second-degree relative in the same or antecedent generations with parkinsonism, but only 1.5% of their relatives at risk had parkinsonism, which is similar to the prevalence in the general population. Ten percent of these patients had at least one relative with essential tremor, but only 1.6% of their relatives at risk had tremor, which again was similar to the prevalence in the population in general. These patients with young onset Parkinson's disease responded well to levodopa therapy. However, dyskinesias and response fluctuations occurred early and frequently. The prevalence of dyskinesias and response fluctuations was strongly correlated with the duration of levodopa treatment, but not with the duration (or probably the severity) of the disease before levodopa therapy was commenced. The involuntary movements often were severe and frequently were diphasic. Despite long disease duration, the incidence of dementia in young onset patients aged less than 65 years was negligible. We believe that most, if not all, patients in this group have degenerative Lewy body idiopathic Parkinson's disease, representing the lower end of a skewed deviation for age of onset of this disease. We have so far failed to identify any additional environmental factor which may have accelerated disease onset in these patients. In contrast, cases of parkinsonism beginning before age 21 years were invariably familial. We proposed that they should be called "juvenile parkinsonism." All affected relatives with parkinsonism also had young disease onset, and all but one were siblings. None of four such patients seen by us has demented, and computed tomography (CT) scan has been normal in all four. We believe that most such patients have some form of genetically determined secondary parkinsonism.