Intra-nasally instilled benzodiazepines have been proposed for acute anxiety episodes. However, routes with faster absorption may increase abuse liability. This study compared abuse liability of intra-nasal midazolam between subjects with a history of intra-nasal drug abuse and non-psychiatric subjects on a single-blind randomized controlled trial. Thirty-one inhaled-cocaine abusers and 34 normal volunteers received either 1 mg intra-nasal midazolam or active placebo. Visual analogue scales assessing desire to repeat the experience (ER) and Experience Liking (EL) assessed abuse liability. Profile analysis for repeated measures showed a significant effect of time over ER (F([5,57])=3.311, p=0.011) and EL (F([5,57])=3.947, p=0.004), diagnostic group (cocaine abusers scoring higher on both - F([5,57])=5.229, p=0.026; F([5,57])=4.946, p=0.030), regardless of the administered substance. It is concluded that the intra-nasal route does not seem to pose risks for non-psychiatric individuals, but it may represent a risk in itself for subjects with a history of drug abuse through this path.