This work evaluates the feasibility of the supercritical enhanced atomization (SEA) process to improve stability and delivery of active pharmaceutical ingredients (APIs). This process was used to generate distinct microcomposites of a model API – theophylline (TPL) – namely pure TPL, theophylline-saccharin (TPL-SAC) co-crystal, and dispersions of each crystalline form in hydrogenated palm oil (HPO) – TPL-HPO and TPL-SAC-HPO. The formation of TPL-SAC co-crystal within the HPO suggests that the co-crystallization step anticipates the lipid dispersion during the formation of the microcomposites. The TPL-SAC co-crystal extended the TPL stability at 92% relative humidity by over 6 months, contrarily to that of raw TPL, which converted into a monohydrate after a few days only, even when dispersed into HPO. The TPL-SAC co-crystal slowed the TPL release from the lipid particles, which is explained by its higher stability towards hydration. The feasibility of the co-crystal microcomposites for therapeutic application was evaluated by estimating the plasmatic concentration of TPL using a one-compartment pharmacokinetic model. The small therapeutic concentration window and high elimination rate of TPL raises serious limitations to control the TPL release. The microcomposites were able to attenuate the TPL burst effect and improve stability towards hydration, but could not extend significantly its delivery.