This thesis investigates the data-driven tracking control design for discrete-time linear systems. Contrary to the traditional model-based design where a mathematical model is identified for the plant and then a controller design follows, the proposed design does not need the step of system modelling. Instead, it can be done directly based on the experimental data collected from the plant. This provides a good alternative to the controller design for those plants that are difficult to identify their models. Data-driven tracking control design via state feedbacks and dynamic output feedbacks are considered. Simulation of several numerical examples and an electronic circuit experiment show the effectiveness of the proposed design.