 We developed a new data-driven distributed control algorithm that uses trajectory data to synthesize a model predictive controller. This algorithm is able to operate with minimal data requirements and can be implemented using distributed computation and local information sharing. We showed that this algorithm has theoretical guarantees for recursive feasibility and asymptotic stability, as well as demonstrated its optimality and scalability in a simulation experiment. This article was authored by Carmen Amo Alonso, Feng Junyang, and Nikolai Matney.