Parkinson’s disease (PD) is a progressive neurodegenerative disorder characterized by motor as well as non-motor symptoms. While available symptomatic therapy is effective, PD is associated with increasing disability and limitations in performance of activities of daily living (ADL). Maintaining and improving activity performance is therefore a major goal in available symptomatic therapy. The most commonly used rating scale for quantification of ADL outcomes is the ADL section (part II) of the Unified PD Rating Scale (UPDRS). However, the UPDRS II does not only include activity but also impairment items. The consequences of this does not appear to have been addressed from a measurement perspective. This paper describes results from Rasch Measurement Theory analyses of complete UPDRS II (full scale, Activity and Impairment) data (n=675-687). Results show poor fit of the full UPDRS II with, e.g., disordered response thresholds, significant misfit and DIF (reliability, 0.82). The 6 Activity items showed no disordered thresholds, significant misfit or DIF (reliability, 0.86), whereas Impairment items did (reliability, 0.51). In conclusion, dividing the UPDRS II into two scales improves ADL measurement, but alternative approaches are needed and the UPDRS II cannot be recommended for measuring ADL in PD as it may yield misleading outcomes.