The present study examined the subjective, psychomotor and reinforcing effects of 10%, 20%, 30% and 40% nitrous oxide in oxygen in 16 healthy volunteers using a choice procedure in which sampling (e.g. 20% nitrous oxide and oxygen-placebo) and choice trials (e.g. 20% nitrous oxide vs. oxygen-placebo) were within the same session. Across the four-session study nitrous oxide dose was varied, Nitrous oxide in a dose-related manner altered subjective effects (e.g. increased visual analog scale ratings of ''high'', ''stimulated'' and ''tingling'') and decreased performance on the Digit Symbol Substitution Test. 10%, 20%, 30% and 40% nitrous oxide were chosen over oxygen by 6, 7, 7 and 8 subjects, respectively. We conclude that nitrous oxide across a range of subanesthetic doses did not function as reinforcer in the majority of subjects tested.