Automated, precise single particle manipulation in the microscale is in great demand and is one of the great challenges in biomedical and biochemical engineering. Automatic micromanipulation has also become a microrobotics challenge. Following this challenge, control technology is integrated with dielectrophoresis (DEP)-based micromanipulation technology in this paper to construct automatic DEP-based micromanipulation systems. DEP micromanipulation systems with electrodes of quadrupole polynomial geometry are developed as controllable microactuators. A semianalytical modelling method is proposed to formulate the analytical models of the DEP manipulation systems, which manifests that the DEP manipulation systems are non-affine non-linear systems. Then, taking the parameter uncertainties, unmodelled dynamics and external disturbances into account, an adaptive law combined with a dynamic sliding mode controller is designed for two-dimensional trajectory tracking control of a DEP micromanipulation system. The closed-loop system is proved stable in the presence of bounded lumped uncertainty based on the Lyapunov theorem. Finally, simulation results show the validity of the proposed control design.