Dual orexin receptor antagonists (DORAs) enable initiation and maintenance of sleep in patients with primary insomnia. Blockade of the orexin system has shown reduction of drug-seeking behavior in animal studies, supporting the role of orexin antagonism as a novel approach for treating substance abuse. Since hypnotics are traditionally associated with misuse, a lack of abuse liability of DORAs would offer significant benefits over current therapies for sleep disorders. In this randomized, crossover, proof-of-concept study, single oral doses of the DORA almorexant (200, 400, and 1,000 mg) were administered to healthy subjects with previous non-therapeutic experience with central nervous system depressants and were compared with placebo and single oral doses of zolpidem (20 and 40 mg), a benzodiazepine-like drug. Subjective measures of abuse potential (visual analog scales [VAS], Addiction Research Center Inventory, and Subjective Drug Value) and objective measures (divided attention [DA]) were evaluated over 24 h post-dose in 33 evaluable subjects. Drug Liking VAS peak effect (E max; primary endpoint) was significantly higher for all doses of almorexant and zolpidem compared with placebo (p<0.001). Almorexant 200 mg showed significantly less 'Drug Liking' than both zolpidem doses (p<0.01), and almorexant 400 mg had smaller effects than zolpidem 20 mg (p<0.05), while almorexant 1,000 mg was not different from either zolpidem dose. Results were similar for other subjective measures, although almorexant generally showed smaller negative and perceptual effects compared with zolpidem. Almorexant also showed less cognitive impairment compared with zolpidem on most DA endpoints. This study in humans investigating single doses of almorexant is the first to explore and show abuse liability of a DORA, a class of compounds that is not only promising for the treatment of sleep disorders, but also of addiction.