The optimum method for evaluating disease progression in Parkinson's disease (PD) has not been established, and this has implications for clinical trials. The majority of previous studies have utilized change on the Unified Parkinson's disease Rating Scale (UPDRS) as an index of progression. However, the UPDRS has not been validated for this purpose. We utilized exploratory factor analysis (EFA) to evaluate the longitudinal properties of the UPDRS as an index of disease progression in PD. Data was derived from a representative cohort of 122 PD patients followed from diagnosis and assessed every 18–24 months for up to 7.9 years. For each subject the rate of change of each item on the UPDRS-3 was calculated and an EFA was performed using this data. Results were compared with those of previously published EFAs in cross-sectional PD cohorts. The UPDRS-3 retains a stable factor structure when used as an index of disease evolution. The 27 items reduced to 6 factors which accounted for 61.0% of the variance in disease progression. A dominant factor was identified which incorporated axial (gait/postural stability) symptoms and signs. Our analysis indicates that the UPDRS captures meaningful aspects of disease progression in PD, and that it is possible to identify symptom/sign complexes which evolve independently of one another. Progression in PD is predominantly characterized by the development of axial symptoms and signs. This result has implications for pathogenesis and should also inform natural history models of PD thereby allowing identification of meaningful outcome measures for clinical trials of disease-modifying therapies.