The morphology of gelatin nanoparticles loaded with three different drugs (Tizanidine hydrochloride, Gatifloxacin and Fluconazole) and their characteristics of entrapment and release from gelatin nanoparticles were investigated by the analysis on nanoparticle size distribution, SEM and FT-IR in this study. The particles were prepared by nanoprecipitation using water and ethanol as a solvent and a nonsolvent, respectively. The exclusion of a crosslinking agent from the procedure led the system to have an irregularly-shaped morphology. Nonetheless, the uncrosslinked case of Gatifloxacin loading generally led to a more homogeneous population of nanoparticles than the uncrosslinked case of Tizanidine hydrochloride loading. No loading was achieved in the case of Fluconazole, whereas both Tizanidine hydrochloride and Gatifloxacin are observed of being capable of being loaded by nanoprecipitation. Tizanidine hydrochloride-loaded, blank and Gatifloxacin-loaded nanoparticles yielded, under crosslinked condition, 59.3, 23.1 and 10.6% of the used dried mass. The crosslinked Tizanidine hydrochloride-loaded particles showed the loading efficiency of 13.8%, which was decreased to 1.1% without crosslinking. A crosslinker such as glutaraldehyde is indispensable to enhance the Tizanidine hydrochloride-loading efficiency. To the contrary, the Gatifloxacin-loading efficiency for crosslinked ones was lower by a factor of 2-3 times than that for uncrosslinked ones. This is due to the carboxylic groups of Gatifloxacin and the aldehyde groups of glutaraldehyde competing with each other during the crosslinking process, to react with the amino groups of gelatin molecules. The loading efficiency of gelatin nanoparticles reported by other investigators greatly varies. Nevertheless, the loading efficiency reported by us is in good agreement with the drug-loading data of gelatin nanoparticles reported by other investigators. The 80% of loaded Tizanidine hydrochloride was released around 15 h after start-up of the release experiment, while the 20% of loaded Gatifloxacin was released more rapidly, as free Gatifloxacin, than the loaded Tizanidine hydrochloride and it showed the trend of sustained slow release during the remaining period of its release experiment. Furthermore, the result of comparative FT-IR analysis is consistent to that of the corresponding drug release study.