The impact of cellular individuality on host-microbe interactions is increasingly appreciated but studying the temporal dynamics of single-cell behavior in this context remains technically challenging. Here we present a microfluidic platform, InfectChip, to trap motile infected cells for high-resolution time-lapse microscopy. This approach allows the direct visualization of all stages of infection, from bacterial uptake to death of the bacterium or host cell, over extended periods of time. We demonstrate the utility of this approach by co-culturing an established host-cell model, Dictyostelium discoideum, with the extracellular pathogen Klebsiella pneumoniae or the intracellular pathogen Mycobacterium marinum. We show that the outcome of such infections is surprisingly heterogeneous, ranging from abortive infection to death of the bacterium or host cell. InfectChip thus provides a simple method to dissect the time-course of host-microbe interactions at the single-cell level, yielding new insights that could not be gleaned from conventional population-based measurements.