The capacity of morphine to suppress natural and precipitated withdrawal was compared in morphine-dependent rhesus monkeys. A similar severity of withdrawal was induced by 14-hr deprivation or precipitated by naloxone, naltrexone, cyclazocine, Win 44,441 or MR 2266. Regardless of the procedure used to induce withdrawal, behavioral signs were completely suppressed by a cumulative dose of 17.5 mg/kg morphine. Thus, an equivalent level of withdrawal induced by reversible antagonists is as sensitive to subsequent morphine administration as is deprivation-induced abstinence. This is in accordance with the theory that vacancy of opiate receptors normally occupied by morphine is related to the level of abstinence observed. In contrast to Win 44,441, however, an equivalent level of withdrawal precipitated by buprenorphine required 175 mg/kg morphine for complete suppression. This is more informative than comparing duration of action; when given as 24-h pretreatments, a high dose of Win 44,441 (10 mg/kg) was only slightly less effective than buprenorphine (3.2 mg/kg) in antagonizing morphine-induced stupor in normal monkeys. Comparison of the ability of morphine to suppress precipitated withdrawal provides evidence of the relative reversibility of antagonists in vivo and demonstrates the extraordinarily stable nature by which buprenorphine acts at opiate receptors.