Standing compared to sitting enhances cognitive performance in healthy subjects. The effect of stance on cognitive performance has been addressed here in patients with Parkinson’s disease (PwPD). We hypothesized that a simple cognitive task would be less enhanced in PwPD by standing with respect to sitting, because of a larger cognitive effort for maintenance of standing posture than in healthy subjects. We recruited 40 subjects (20 PwPD and 20 age-matched healthy subjects, HE). Each participant performed an arithmetic task (backward counting aloud by 7) in two postural states, sitting and standing, with eyes open (EO) and with eyes closed (EC). All trials lasted 60 s and were randomized across subjects and conditions. The number of correct subtractions per trial was an index of counting efficiency and the ratio of correct subtractions to total subtractions was an index of accuracy. All conditions collapsed, the efficiency of the cognitive task was significantly lower in PwPD than HE, whilst accuracy was affected to a lower extent. Efficiency significantly improved from sitting to standing in HE under both visual conditions whilst only with EO in PwPD. Accuracy was not affected by posture or vision in either group. We suggest that standing, compared to sitting, increases arousal, thus improving the cognitive performance in HE. Conversely, in PwPD this improvement was present only with vision, possibly due to their greater balance impairment with EC consuming an excess of attentional resources. These findings have implications for balance control and the risk of falling in PwPD in the absence of visual cues.