The objective of this study was to investigate the effects of sample preparation, HPLC conditions and peak measurement methods upon determining progesterone content of poly-d,l-lactide-co-glycolide microspheres. A series of the microspheres with different formulations was first prepared. To determine their actual drug contents, the microspheres were dissolved in tetrahydrofuran and diluted with various amounts of methanol to precipitate the polymer. After removal of polymeric precipitates, the filtrates were subject to HPLC analysis under versatile experimental conditions. Interestingly, the composition of a sample solution (e.g., the ratio of methanol to tetrahydrofuran) affected the magnitudes of both peak fronting and peak broadening of progesterone. Its peak became broader and more asymmetrical at lower methanol:tetrahydrofuran ratios. Furthermore, its peak height was influenced by the proportion of tetrahydrofuran in a sample solution. Such problems encountered with tetrahydrofuran were exacerbated when a larger volume of the sample solution was injected onto an analytical column. Under our experimental conditions a peak area measurement provided more accurate and reliable determination of progesterone content in various microspheres than a peak height determination. Optimizing the composition of a sample solution, HPLC chromatographic conditions and peak analysis methods was a prerequisite to an accurate determination of progesterone encapsulated within microspheres.