OBJECTIVE: To develop a generic decision-analytic model for the evaluation of long-term clinical and economic consequences of interventions in patients with Parkinson's disease (PD) which can be applied to different research questions, interventions, and outcomes, and is based on (untreated) biological progression. METHODS: We developed a Markov model, in which a hypothetical cohort of patients moves through health states reflecting patient characteristics that are observed under treatment (Hoehn and Yahr "on" state [HYon]) and would be observed in the absence of treatment (Hoehn and Yahr "off" state [HYoff]). We used HYoff I-V s as Markov states, because those reflect underlying biologic progression of PD. Interventions: diagnostic or treatment strategies in PD patients such as Levodopa, dopamine agonists, or other anti-parkinsonian drugs as well as surgical therapies such as deep brain stimulation. Data: Transition probabilities for HYoff states were derived from the literature. The distribution of HYon was modeled conditional on HYoff using studies that report both characteristic. Complications were modeled conditional on HYon using data from a registry established by the "Competence Network Parkinson Disease". Utilities, and costs were modeled conditional on HYon and presence of complication. Mortality is a function of age- and gender-specific background mortality and PD-specific mortality Time horizon: lifetime with annual cycle length. RESULTS: remaining (quality-adjusted) life expectancy, direct costs, and incremental cost-effectiveness ratios. Further outcomes comprise clinical events or complications such as motor complications, dementia, depression, and hallucinations. In addition, complication-free survival, time in HYoff and HYon states, and UPDRS scores were modeled as additional outcomes. Perspective: societal and third party payer. Sensitivity analysis: 1-way and multiway. An interactive interface allows to model different settings or countries. CONCLUSIONS: In contrast to formerly published models, this generic PD model has the ability to consider multiple interventions and outcomes and to switch between outcomes depending on which outcomes are reported in a clinical trial.