Establishment of in vitro-in vivo correlations (IVIVCs) for parenteral polymeric microspheres has been very challenging, due to their complex multiphase release characteristics (which is affected by the nature of the drug) as well as the lack of compendial in vitro release testing methods. Previously, a Level A correlation has been established and validated for polymeric microspheres containing risperidone (a practically water insoluble small molecule drug). The objectives of the present study were: 1) to investigate whether a Level A IVIVC can be established for polymeric microspheres containing another small molecule drug with different solubility profiles compared to risperidone; and 2) to determine whether release characteristic differences (bi-phasic vs tri-phasic) between microspheres can affect the development and predictability of IVIVCs. Naltrexone was chosen as the model drug. Three compositionally equivalent formulations of naltrexone microspheres with different release characteristics were prepared using different manufacturing processes. The critical physicochemical properties (such as drug loading, particle size, porosity, and morphology) as well as the in vitro release characteristics of the prepared naltrexone microspheres and the reference-listed drug (Vivitrol®) were determined. The pharmacokinetics of the naltrexone microspheres were investigated using a rabbit model. The obtained pharmacokinetic profiles were deconvoluted using the Loo-Riegelman method, and compared with the in vitro release profiles of the naltrexone microspheres obtained using USP apparatus 4. Level A IVIVCs were established and validated for predictability. The results demonstrated that the developed USP 4 method was capable of detecting manufacturing process related performance changes, and most importantly, predicting the in vivo performance of naltrexone microspheres in the investigated animal model. A critical difference between naltrexone and risperidone loaded microspheres is their respective bi-phasic and tri-phasic release profiles with varying burst release and lag phase. These variations in release profiles affect the development of IVIVCs. Nevertheless, IVIVCs have been established and validated for polymeric microspheres with different release characteristics.