As L-Dopa was marketed in France in 1971 for treatment of Parkinson's disease we used 1971 to divide a sample of patients into 2 groups. Group 1 (152 patients) includes patients with diagnosis made before 1971 and group 2 (264 patients) with diagnosis after 1971. The prognostic factors were motor deterioration, intellectual deterioration and death. The prognostic variables include the neuro-psychological status at the onset of the disease and at the beginning of L-Dopa treatment. The statistical analysis is based on Kaplan-Meier estimate, Log-rank test and Cox's model. In group 2, 10 years after the beginning treatment of L-Dopa, motor deterioration affected 60 p. 100 of the patients. The poor variables were akineto-hypertonic type, severe akinesia, poor clinical result after one year. Intellectual deterioration was frequent: 30 p. 100 at 10 years. The poor variables were age over 60, depression or psychotic episodes occurring during the first year. The 10 years-survival rate was 64 p. 100 and was not different from that recorded in a French population of same age and sex distribution. The poor variables were severe akinesia, presence of a Babinski sign, poor therapeutic result after one year, occurrence of psychotic episodes during the first year. The interval between the onset of the disease and the beginning of L-Dopa did not have any prognostic value, whatever the response criterion. In group 1, the same prognostic factors were pointed out. Survival and intellectual deterioration were not different in the 2 groups, but in group 1, motor deterioration appeared earlier.(ABSTRACT TRUNCATED AT 250 WORDS)