OBJECTIVES: Dyskinesia is generally thought to severely affect health-related quality of life (QoL) in patients with Parkinson’s disease (PD). We evaluated the independent impact of dyskinesia on QoL in patients with PD and evaluated whether commonly used QoL instruments are sensitive enough to measure dyskinesia effects in clinical trials. METHODS: We analyzed data from the German PD Competence Network comprising generic (EuroQoL [EQ-5D]) and disease-specific (PD Questionnaire 39 [PDQ-39]) QoL instruments and clinical variables including the Unified Parkinson’s Disease Rating Scale (UPDRS). We used 4 dyskinesia-specific UPDRS items (i.e., duration, disability, painfulness of dyskinesias and presence of early-morning dystonia) to predict totals and subscores of EQ-5D and PDQ-39, values of the visual analogue scale (VAS) and EQ-5D derived utilities. We performed ordinal logistic regression to predict EQ-5D subscales and multiple linear regression to predict all remaining QoL outcomes. Potential confounders were specified a priori by an expert panel and final confounders were selected based on statistical criteria (univariate Spearman’s rank correlation, multivariate forward selection, p<0.05). RESULTS: A total of 68 models were investigated (4 dyskinesia x 17 QoL variables), of which 9 showed a statistically significant association after controlling for confounding. The most relevant confounder was severity of disease. All 4 dyskinesia variables were associated with at least 1 QoL variable. Dyskinesia duration was shown to be the most robust predictor. Subscales of EQ-5D and PDQ-39 addressing pain/(bodily) discomfort were associated with all 4 dyskinesia variables. In addition, EQ-5D index was associated with duration of dyskinesia. CONCLUSIONS: Dyskinesia has a significant impact on QoL measured by EQ-5D and PDQ-39 and their subscales, even after controlling for confounding. EQ-5D and PDQ-39 are useful instruments for clinical trials addressing questions on the effect of potential anti-dyskinesia treatments on QoL.