Ketoprofen was encapsulated within poly-epsilon-caprolactone (PCL) and hydroxypropyl methylcellulose phthalate 50 (HPMCP50) microspheres (MS). Scanning electron microscopy (SEM) studies showed spherical particles without surface crystal formation and differential scanning calorimetry (DSC) supported these results. MS of PCL or HPMCP50 had a mean particle size of 10.7 +/- 2.2 and 10.9 +/- 2.0 mu m respectively, whereas a mixture of these polymers increased the MS particle size to 30 mu m. Greater incorporation efficiencies were found for HPMCP50 MS (98.1 +/- 0.7%). MS of PCL and HPMCP50 mixtures showed a decreased drug entrapment as the amount of PCL was increased (96.0 +/- 0.2 for 25% PCL, 95.6 +/- 1.8 for 50% PCL, 80.2 +/-0.7 for 75% PCL and 78.9 +/- 9.0 for 100% PCL). Size exclusion chromatography (SEC) studies revealed a weak interaction between ketoprofen and PCL and some polymer degradation was found during HPMCP50 MS storage, probably by breaking of the phthalic anhydride bond to be anyhydroglucose backbone. Four types of cryoprotectors (glucose, trehalose, mannitol and sorbitol, at 5 and 10% W/V) and two freezing conditions (-196 and -20 degrees C) were evaluated in freeze-drying studies. For HPMCP50, the sizes of MS after reconstitution of liophylizates were nearly the same as the initial ones. For PCL MS only, those formulations with sorbitol or glucose at 10% and frozen at -196 degrees C showed acceptable results. In contrast to the rapid release rate of ketoprofen from PCL MS as a result of carrier porosity (80% released within 15 min), the release from HPMCP50 MS could be controlled by means of pH (40% released in the first 15 min in simulated gastric fluid and nearly 100% ketoprofen delivered in the same time in simulated intestinal fluid).