Preclinical evidence suggests that ghrelin, a peptide synthesized by endocrine cells of the stomach and a key component of the gut–brain axis, is involved in alcohol seeking as it modulates both central reward and stress pathways. However, whether and how ghrelin administration may impact alcohol intake in humans is not clear. For, we believe, the first time, this was investigated in the present randomized, crossover, double-blind, placebo-controlled, human laboratory study. Participants were non-treatment-seeking alcohol-dependent heavy-drinking individuals. A 10-min loading dose of intravenous ghrelin/placebo (3 mcg kg− 1) followed by a continuous ghrelin/placebo infusion (16.9 ng/kg/min) was administered. During a progressive-ratio alcohol self-administration experiment, participants could press a button to receive intravenous alcohol using the Computerized Alcohol Infusion System. In another experiment, brain functional magnetic resonance imaging was conducted while participants performed a task to gain points for alcohol, food or no reward. Results showed that intravenous ghrelin, compared to placebo, significantly increased the number of alcohol infusions self-administered (percent change: 24.97±10.65, P= 0.04, Cohen’s d= 0.74). Participants were also significantly faster to initiate alcohol self-administration when they received ghrelin, compared to placebo (P= 0.03). The relationships between breath alcohol concentration and subjective effects of alcohol were also moderated by ghrelin administration. Neuroimaging data showed that ghrelin increased the alcohol-related signal in the amygdala (P= 0.01) and modulated the food-related signal in the medial orbitofrontal cortex (P= 0.01) and nucleus accumbens (P= 0.08). These data indicate that ghrelin signaling affects alcohol seeking in humans and should be further investigated as a promising target for developing novel medications for alcohol use disorder.