Tensegrity structures have been widely utilized as lightweight structures due to their high stiffness-to-mass and strength-to-mass ratios. Minimal mass design of tensegrity structures subject to external loads and specific constraints (e.g., member yielding and buckling) has been intensively studied. However, all the existing studies focus on passive tensegrity structures, i.e., the structural members cannot change their lengths actively and the structure has to passively resist external loads. An active tensegrity structure equipped with actuators can actively adapt its internal forces and nodal positions and thus can actively resist external loads. Therefore, it is expected that active tensegrity structures use less material compared to passive tensegrity structures thus leading to a smaller mass. Due to the integration of the active control system, the design of active tensegrity structures is different from passive tensegrity structures. This study proposes a general approach for the design of minimal mass active tensegrity structures based on a mixed integer programming scheme, in which the member crosssectional areas, prestress, actuator layout and control strategies (i.e., actuator length changes) are designed simultaneously. The member cross-sectional areas, prestress level, and actuator control strategies are treated as continuous variables and the actuator layout is treated as a binary variable. The equilibrium condition, member yielding, cable slackness, strut buckling, and the limitations on the nodal displacements as well as other practical requirements are formulated as constraints. Three typical active tensegrity structures are designed through the proposed approach and the results are benchmarked with the equivalent minimal mass passive designs. It is illustrated that the active designs can significantly decrease the material consumption compared with the equivalent passive designs thus leading to more lightweight tensegrity structures.