Nanoparticulate drug delivery systems have several opportunities to overcome bioavailability or stability problems of peptide and protein drugs. In this study a 23 full factorial design was used for preparation of insulin containing nanoparticles using different concentrations of low, medium and high molecular weight chitosan (CS) and tripolyphosphate (TPP) by ion gelation method. Encapsulation efficiencies (EE) of each formulation were determined by HPLC method. Regression analysis and surface plots were used in order to evaluating the effect of variables on EE and choosing the optimum formulations. The morphology of selected nanoparticles was obtained by transmission electron microscopy (TEM). Particle size, poly dispersity index (PDI) and zeta potential were also measured. Freeze-dried nanoparticles were used for drug release studies in phosphate buffer (pH=6.8). Resulted nanoparticles had mean size of 112-419 nm with PDI< 0.5 and positive zeta potential. Insulin concentration and molecular weight of chitosan had pronounced effect on EE, but chitosan concentration had no considerable effect on EE. The maximum EE of CS nanoparticles with low, medium and high molecular weights were 61.88 %, 70.89 % and 53.73 %, respectively. The in vitro drug release profiles from the low molecular weight chitosan nanoparticles showed an initial burst release followed by a slow release within 3 hours.