The risk of developing levodopa-associated dyskinesia is known to vary inversely with the age of Parkinson's disease onset. This study quantifies dyskinesia risks for different Parkinson's onset ages in a patient population treated at the Parkinson's Disease Research, Education, and Clinical Center at the San Francisco Veterans Affairs Medical Center. Medical records were reviewed to determine age of Parkinson's onset, medication history, and dyskinesia onset. Dyskinesia risks were determined by using Kaplan-Meier analysis. Cox proportional hazard models were used to compare age groups and to perform multivariate modeling. This study included 109 patients with Parkinson's, 105 of whom had onset of symptoms after 1989. At 5 years of levodopa treatment, the dyskinesia risk for patients with onset age 40-49 was 70%, decreasing to 42% for onset ages 50-59, 33% for onset ages 60-69, and 24% for onset ages 70-79. Pairwise comparisons between the 40-49 age group and the other age groups were statistically significant in time-to-event models. After 5 years of levodopa treatment, dyskinesia risks became uniformly high regardless of age of onset. These results suggest it is appropriate to use different baseline dyskinesia risks in clinical decision-making for patients on the basis of their ages of onset. However, the most significant difference occurs between ages 40-49 and ages 50-79, and if more than 5 years of levodopa therapy are anticipated, dyskinesia risk may have less utility when deciding upon Parkinson's therapy. Drug studies for Parkinson's disease should also take age of Parkinson's onset into account when analyzing dyskinesia outcomes.