Previously, we reported that gabapentin, a lambda-aminobutyric acid (GABA) agonist, significantly reduced "positive" subjective effects of cocaine without reducing cocaine self-administration. We speculated that the gabapentin doses used in that study were too low to detect subtle shifts in the reinforcing effects of cocaine. Thus, the present study examined the effects of larger gabapentin maintenance doses on cocaine-related effects, including self-administration. During this 48-day double-blind, crossover design study, the effects of gabapentin maintenance (0, 2400, 3200 mg/day) on response to cocaine (0, 12, 25, 50 mg) were investigated in six cocaine-dependent individuals not seeking treatment for their cocaine use. Active cocaine significantly increased choice to self-administer cocaine, subjective-effect ratings (e.g., "Good Drug Effect"), blood pressure, and heart rate. Gabapentin did not decrease cocaine self-administration, cardiovascular measures, or most subjective effects of cocaine. These data agree with findings from a clinical trial examining the effects of similar gabapentin doses on cocaine use by treatment-seeking cocaine-dependent individuals and suggest that gabapentin does not show promise as a treatment medication for cocaine dependence.