OBJECTIVES: To investigate the relationships between freezing of gait (FOG) and activity limitations in ambulant people with Parkinson's disease (PD), and to explore the contribution of FOG and gait hypokinesia to activity limitations after adjusting for the effects of disease severity. DESIGN: Retrospective, cross-sectional design. SETTING: Participants were recruited from neurologists' clinics and the general public in metropolitan Melbourne, Australia. PARTICIPANTS: Volunteers were screened for eligibility using the following inclusion criteria: diagnosis of idiopathic PD, modified Hoehn and Yahr stages 0 to IV, without dementia. Participants (N=210) were tested (mean age ± SD, 67.9±9.6y; mean PD duration ± SD, 6.7±5.6y; median Hoehn and Yahr stage=2.5). INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: FOG was measured using the FOG questionnaire. Gait hypokinesia was quantified using both the 6-meter walk test and the Timed Up and Go test. Activity limitation was measured using the modified Unified Parkinson's Disease Rating Scale activities of daily living (ADL) section and the Schwab and England ADL scale. RESULTS: Severity of gait freezing correlated significantly with the level of activity limitation (Spearman correlation coefficient, ρ=-.49 to .48; P<.001). A hierarchic regression model showed that disease severity explained 37.5% of the variance in Schwab and England ADL score (P<.001). Gait hypokinesia and FOG severity scores explained an additional 9.1% of the variance in activity limitation (R(2) change=.091; P<.001). CONCLUSIONS: FOG severity and gait hypokinesia were associated with reduced levels of activity after adjusting for disease severity.