Morphine-dependent and control rats in an oral free-choice protocol were treated with gamma-vinyl GABA (GVG), 60, 120 and 240 mg/kg IP, for 3 days over three successive periods. Morphine dependence was assessed with naltrexone. Fluid intake of morphine-dependent rats was reduced during GVG treatment and the proportion of fluid taken as morphine was also decreased, but not sufficiently to induce withdrawal signs. Intake by controls was only affected by 240 mg/kg GVG, when body weight decreased. The attentuation of morphine consumption by GVG treatment may, within the limitations of this protocol, interfere with the positive reinforcing properties of morphine.