The purpose of the present study was to characterize the in vivo release kinetics of octreotide acetate from microsphere formulations designed to minimize peptide acylation and improve drug stability. Microspheres were prepared by a conventional oil/water (o/w) method or an experimental oil/oil (o/o) dispersion technique. The dosage forms were administered subcutaneously to a rat animal model, and serum samples were analyzed by radioimmunoassay over a 2-month period. An averaged kinetic profile from each treatment group, as a result, was treated with fractional differential equations. The results indicated that poly(l-lactide) microspheres prepared by the o/o dispersion technique provided lower area under the curve (AUC) values during the initial diffusion-controlled release phase, 7.79 ngxd/mL, versus 75.8 ngxd/mL for the o/w batch. During the subsequent erosion-controlled release phase, on the other hand, the o/o technique yielded higher AUC values, 123 ngxd/mL, versus 42.2 ngxd/mL for the o/w batch. The differences observed between the 2 techniques were attributed to the site of drug incorporation during the manufacturing process, given that microspheres contain both porous hydrophilic channels and dense hydrophobic matrix regions. An o/o dispersion technique was therefore expected to produce microspheres with lower incorporation in the aqueous channels, which are responsible for diffusion-mediated drug release.