Studies in rodents suggest that insulin controls hepatic glucose metabolism through brain-liver crosstalk, but human studies using intranasal insulin to mimic central insulin delivery provided conflicting results. In this randomized controlled cross-over trial we investigated effects of intranasal insulin on hepatic insulin sensitivity and energy metabolism in 10 patients with type 2 diabetes and 10 lean healthy participants (CON). Endogenous glucose production (EGP) was monitored with [6,6-(2)H2]glucose, hepatocellular lipids (HCL), adenosine triphosphate (ATP) and inorganic phosphate concentrations with (31)P/(1)H magnetic resonance spectroscopy. Intranasal insulin transiently increased serum insulin, followed by gradual lowering of blood glucose in CON only. Fasting hepatic insulin sensitivity index (HIS) was not affected by intranasal insulin in CON and patients. Only in CON, HCL decreased by 35%, whereas absolute hepatic ATP concentrations increased by 18% after three hours. A subgroup of CON received intravenous insulin to mimic the changes in serum insulin and blood glucose levels observed after intranasal insulin. This resulted in a 34% increase in HCL, without altering hepatic ATP concentrations. In conclusion, intranasal insulin does not affect HIS but rapidly improves hepatic energy metabolism in healthy humans, which is independent of peripheral insulinemia. These effects are blunted in patients with type 2 diabetes.