A method designed to quantitate in vivo naltrexone release rates from sustained-release systems has been applied to the evaluation of seven different naltrexone delivery systems in the monkey. The method consists of two phases: a single intravenous bolus dose quantitation of each monkey's pharmacokinetic parameters coupled with a delivery system study in which plasma naltrexone levels are measured throughout the time period of sustained-release. In vivo release rates and the total amount released are then calculated. It should be noted that these determinations require the analysis of unchanged naltrexone in plasma as the only experimental measurement. Data from injectable naltrexone pamoate microcapsule delivery systems indicate that 1) when these microcapsules are suspended in an aqueous vehicle, a significant part of the dose is released very rapidly, yielding release rate-time data that parallel a non-sustained-release control; 2) this rapid release for the aqueous vehicle is followed by a slow release phase lasting to about 24 days for the subcutaneous route and to about 45 days for the intramuscular route; and 3) when these microcapsules are suspended in an oily vehicle there is no initial rapid release, substantial release rates are obtained for at least 60 days, and an average of 89% of the dose is calculated to have been released. Data from implantable naltrexone delivery systems show that 1) the Alza system most closely approximates a zero-order release rate-time profile; 2) the Battelle system provides a rapid initial release followed by a slowly declining release rate; 3) the Dynatech system is characterized by a more rapid initial release rate of 3-8% of the dose per day over the first 3-5 days followed by a rather constant 1-3% per day to about day 36; and 4) essentially complete recovery of the dose was obtained for the Battelle and Dynatech systems.