The tracking of cell populations in time-lapse microscopy images enables high-throughput spatiotemporal measurements of cell dynamics. In this paper, we present a new algorithm to simultaneously track many cells in crowded areas. The algorithm runs in real time and deals with thousands of cells. The main contribution of this paper is that the algorithm is able to maintain the spatiotemporal consistency of the tracks in crowded areas, even when the temporal resolution is coarse. We validate our approach in terms of its ability to track yeast cells.