Microencapsulation has been a promising approach for drug delivery, cell implantation, cell-based gene therapy and large-scale cell culture. To make use of microcapsules more effectively, it is important to accurately construct the microcapsule membranes with desired properties including a certain thickness, strength, and so forth. To date single factor experiments have been widely used, however, they are time-consuming to obtain the desired membrane preparation conditions. Response surface methodology (RSM) is a mathematical and statistical technique for building empirical models that gained importance for optimizing reacting conditions. In this study, three signifficant effect factors that affect alginate-based microcapsule membrane properties, including membrane thickness, swelling degree, and mechanical stability, were determined with Plackett-Burman method, and then three empirical models were built to optimize the preparation conditions of the microcapsule membranes according to the responses of these three signifficant effect factors respectively with RSM. These models can be used to predict the characteristics of microcapsules under different membrane preparation conditions, which provide a guide for optimizing the microencapsulation technology.