Key points: Preclinical models have demonstrated that nitric oxide is a key component of neurovascular coupling; this has yet to be translated to humans. We conducted two separate protocols utilizing intravenous infusion of a nitric oxide synthase inhibitor and isovolumic haemodilution to assess the influence of nitric oxide on neurovascular coupling in humans. Isovolumic haemodilution did not alter neurovascular coupling. Intravenous infusion of a nitric oxide synthase inhibitor reduced the neurovascular coupling response by ∼30%, indicating that nitric oxide is integral to neurovascular coupling in humans. Abstract: Nitric oxide is a vital neurovascular signalling molecule in preclinical models, yet the mechanisms underlying neurovascular coupling (NVC) in humans have yet to be elucidated. To investigate the contribution of nitric oxide to NVC in humans, we utilized a visual stimulus paradigm to elicit an NVC response in the posterior cerebral circulation. Two distinct mechanistic interventions were conducted on young healthy males: (1) NVC was assessed during intravenous infusion of saline (placebo) and the non-selective competitive nitric oxide synthase inhibitor NG-monomethyl-l-arginine (l-NMMA, 5 mg kg−1 bolus & subsequent 50 μg kg−1 min−1 maintenance dose; n = 10). The order of infusion was randomized, counterbalanced and single blinded. A subset of participants in this study (n = 4) underwent a separate intervention with phenylephrine infusion to independently consider the influence of blood pressure changes on NVC (0.1–0.6 μg kg−1 min−1 constant infusion). (2) NVC was assessed prior to and following isovolumic haemodilution, whereby 20% of whole blood was removed and replaced with 5% human serum albumin to reduce haemoglobin concentration (n = 8). For both protocols, arterial and internal jugular venous blood samples were collected at rest and coupled with volumetric measures of cerebral blood flow (duplex ultrasound) to quantify resting cerebral metabolic parameters. l-NMMA elicited a 30% reduction in the peak (P = 0.01), but not average (P = 0.11), NVC response. Neither phenylephrine nor haemodilution influenced NVC. Nitric oxide signalling is integral to NVC in humans, providing a new direction for research into pharmacological treatment of humans with dementia.