Small animals such as the roundworm C. elegans are excellent models for studying bacterial infection and host response, as well as for genetic and chemical screens. A key methodology is the killing assay, in which the number of surviving animals is tracked as a function of the time post infection. This is a labor-intensive procedure, prone to human error and subjective choices, and often involves undesired perturbation to the animals and their environment. In addition, the survival of animals is just one aspect of a multi-dimensional complex biological process. Here we report a microfluidic-based approach for performing killing assays in worms, compatible with standard assays performed on solid media. In addition to providing accurate and reproducible survival curves at a considerably reduced labor, this approach allows acquisition of a multitude of quantitative data with minimal undesired perturbations. These measurements are obtained automatically at a worm-by-worm resolution using a custom image processing workflow. The proposed approach is simple, scalable, and extendable, and is significantly more economical than standard manual protocols.