Support vector machine (SVM), which is a novel technology for solving classification and regression issues, has been successfully used in many fields. In this study, according to an experimental dataset on the average grain size of Fe 3 O 4 nanoparticles, a predicting and optimizing model using support vector regression (SVR) was developed. In this model, the estimated result of SVR agreed with the experimental data well. In addition, the particle swarm optimization (PSO) algorithm is employed for optimizing the parameters of SVR models and obtaining the optimal process parameters for preparing Fe 3 O 4 nanoparticles. The minimum grain size of Fe 3 O 4 nanoparticles is forecasted to be 10nm while the Fe 3 O 4 nanoparticles are synthesized by using the optimal process parameters. Meanwhile, multifactor analysis is conducted for investigating the influence of process parameters on the average grain size of Fe 3 O 4 nanoparticles. The results suggest that SVR is capable of providing important theoretical and practical guide in research and development of Fe 3 O 4 nanoparticles possessing ideal grain size.