Digital labour platforms (DLPs) digitally connect human workers with consumers, for a type of work known as digital platform work. DLPs face challenges due to the complexity of control of digital platform work. Accordingly, they execute control through algorithmic means, which, however, presents problematic scenarios, such as lack of autonomy and fairness, and low worker well-being. Problematizing control executed solely through algorithms, we develop the concept of control partitioning, theorizing the distribution of the control of digital platform work between multiple controllers (i.e., DLPs, third-party intermediaries, and customers) and between algorithmic and human control mechanisms. Control partitioning provides a novel theoretical perspective that brings the human back into the loop of control, in contrast to existing studies which suggest DLPs as the only controllers and algorithmic control as the exclusive control means.