Distance/time measures of gait in 105 sufferers from idiopathic Parkinsonism, who were able to walk unaided, and 144 healthy controls were examined systematically. Those sufferers with overt fluctuations in control were assessed during their "therapeutic window". Free walking speed was lower for a given cadence in the sufferers, but reached a plateau whilst cadence could still be increased. Age, cognitive function and the range of passive hip flexion were important determinants of gait in them. Even minor degrees of cognitive impairment were associated with reduced free walking speed in sufferers: it appears unwise that they were prescribed more sedatives than the controls. The potential benefit of physiotherapy in maintaining joint flexibility was noted. The deficits in speed of individual sufferers, and hence the estimated potential for prophylaxis and treatment, were unrelated to age at presentation. There was no evidence for a limited period of responsiveness to levodopa therapy in this cross-sectional study.