The aim of this work was to study cerebral vasoreactivity to hypercapnia in Parkinson's disease (PD) before and after levodopa administration. The prospective study was conducted in 20 patients presenting with PD, using 3T blood oxygenation level-dependent (BOLD) functional MRI (fMRI) covering the whole brain. The hypercapnic stimulus was block-designed using carbogen inhalation, a gas mixture of 7% CO2 and 93% O2, before (OFF) and 60 minutes after administration of a suprathreshold (120%) therapeutic L-dopa dose (ON). Ten age-matched controls were enrolled for between-group comparisons. Analyses were conducted with a random effects model and corrected for multiple comparisons. No adverse reaction to the hypercapnic stimulus was reported. However, 10 patients and 2 controls were excluded because of incomplete protocol realization, inappropriate hypercapnic stimulus, or excessive movements, leaving 10 patients and 8 controls for further analyses. The hypercapnic stimulus increased whole-brain BOLD signal of 1.48% ± 0.06% (mean ± standard error) in controls, 1.59% ± 0.05% in patients OFF, and 1.62% ± 0.09% in patients ON. Regions of interest analyses showed a signal increase in gray matter of 2.60% ± 0.16% in controls, 2.89% ± 0.21% in patients OFF, and 2.87% ± 0.12% in patients ON. No global or regional significant difference was detected, when comparing patients OFF and ON L-dopa, or between patients and controls. Contrary to Alzheimer's disease, the vasoreactivity to hypercapnia was normal in PD before and after L-dopa administration, compared to controls. This negative result is an important finding, especially for neuroscientists using fMRI to investigate motricity and cognition, discarding a significant confounding effect.