Three opioid antagonists (MR2266, 16-methyl cyprenorphine and nor-binaltorphimine) were tested independently for their ability to suppress the intake of a highly palatable saccharin and glucose (S/G) solution after central administration. MR2266 is an equally potent antagonist at kappa (κ) and mu (μ) opioid receptors. Nor-binaltorphimine (N-BNI) and 16-methyl cyprenorphine (M80) are two recently developed opioid antagonists that were chosen based upon their ability to act more selectively than naloxone at κ and delta (σ) opioid receptor types, respectively. Prior research has demonstrated that when dissolved in acid and administered centrally, MR2266 (20 μg) fails to suppress S/G intake. Because all three antagonists are rather insoluble in water, they were dissolved in dimethyl sulfoxide (DMSO). Rats with chronic ventricular cannula were allowed to consume S/G for a 0.5 hr bout. They received a single intracerebroventricular (ICV) injection of antagonist (MR2266: 0, 10, 20 and 40 μg; M80: 0, 5, 10, 20 and 40 μg or N-BNI: 0, 1, 3, and 10 μg) 10 min prior to the start of the drinking bout. Administration of DMSO alone failed to alter drinking relative to saline, whereas each antagonist significantly attenuated S/G intake. We conclude that, when dissolved in DMSO, these antagonists suppress drinking by blockade of opioid receptors. © 1990.