This thesis examines the principle of purpose limitation in data protection law from the perspective of regulating data-driven innovation. According to this approach, the principle of purpose limitation not only protects an individual's autonomy but simultaneously leaves sufficient room for data controllers to innovate when finding the best solution for protection. The first component of the principle of purpose limitation (i.e. to specify the purpose of data processing) is a precautionary protection instrument which obliges the controller to identify specific risks arising from its processing against all fundamental rights of the data subject. In contrast, the second component (i.e. the requirement to limit data processing to the preceding purpose) aims to control the risk caused by data processing that occurred at a later stage and adds to the risks which were previously identified. This approach provides an answer to the question of how the General Data Protection Regulation which does not only effectively protect an individual's autonomy but also helps controllers to turn their legal compliance into a mechanism that enhances innovation, should be interpreted with regard to all the fundamental rights of the data subject.