This study aimed to characterise three nasal drug delivery devices to evaluate their propensity to deliver human insulin solutions to the nasal cavity for redistribution to the central nervous system. Brain delivery was evaluated using functional magnetic resonance imaging to measure regional cerebral blood flow. Intranasal insulin administration has been hypothesised to exploit nose-to-brain pathways and deliver drug directly to the brain tissue whilst limiting systemic exposure. Three nasal pump-actuator configurations were compared for delivery of 400 IU/mL insulin solution by measuring droplet size distribution, plume geometry, spray pattern and in vitro deposition in a nasal cast. The device with optimal spray properties for nose to brain delivery (spray angle between 30° and 45°; droplet size between 20 and 50 μm) also favoured high posterior-superior deposition in the nasal cast and was utilised in a pharmacological magnetic resonance imaging study. Functional magnetic resonance imaging in healthy male volunteers showed statistically significant decreases in regional cerebral blood flow within areas dense in insulin receptors (bilateral amygdala) in response to intranasally administered insulin (160 IU) compared to saline (control). These changes correspond to the expected effects of insulin in the brain and were achieved using a simple nasal spray device and solution formulation. We recommend that a thorough characterisation of nasal delivery devices and qualitative/quantitative assessment of the administered dose is reported in all studies of nose to brain delivery so that responses can be evaluated with respect to posology and comparison between studies is facilitated.