OBJECTIVE： To study release behavior in vitro of Risperidone sustained-release tablets and its pharmacokinetics in rabbits. METHODS： Risperidone sustained-release tablets were prepared by using mesoporous silica as matrix. Release rates in vitro within 12 h （Q12 h） of commercially available Risperidone tablets， Risperidone sustained-release tablets and its physical mixture in 0.1 mol/L HCl fluid were investigated with basket method. The release model of Risperidone sustained-release tablets were fitted. Using clozapine as an internal standard， HPLC method was used to determine blood concentration of risperidone and 9-hydroxyrisperidone in rabbits 48 h （n＝6） after intragastric administration of commercially available Risperidone tablets and Risperidone sustained-release tablets 2 mg. Pharmacokinetic parameters were calculated by using non-compartmental model of Kinetica 4.4 software. RESULTS： Compared with commercially available Risperidone tablets （Q12 h＝97％） and physical mixtures （Q12 h＝95％）， release rate of Risperidone sustained-release tablets （Q12 h＝83.7％） slowed down significantly， and the release of Risperidone sustained-release tablets in 0.1 mol/L HCl fluid was closed to first-order release （R2＝0.998 9）， with diffusion as the main factor and dissolution as the supplement. By risperidone， the pharmacokinetic parameters of commercially available Risperidone tablets and Risperidone sustained-release tablets included that t1/2 were （4.64±0.93），（6.65±0.92） h； cmax were （34.46±7.75） and （8.57±6.91） ng/mL； MRT were （11.48±1.23）， （17.46±2.10） h； AUC0-48 h were （314.39±10.33），（192.98±49.14） ng·h/mL， respectively. By 9-hydroxyrisperidone， the pharmacokinetic parameters of them included that t1/2 were（7.08±0.93），（10.45±0.78） h； cmax were （98.08±5.43），（54.55±4.88） ng/mL； MRT were （11.48±1.23）， （17.46±2.10） h； AUC0-48 h were （894.71±131.15）， （1 227.99±112.12） ng·h/mL （n＝6）， respectively. Compared with commercially available Risperidone tablets， t1/2 and MRT of Risperidone sustained-release tablets prolonged significantly， while cmax decreased significantly （P＜0.05）. CONCLUSIONS： Risperidone loaded in mesoporous silica has sustained release effect and prolong the time of drug efficacy.