Tuberculosis is considered to be one of the world's deadliest disease with 2 million deaths each year. The need for new antitubercular drugs is further exacerbated by the emergence of drug-resistance strains. Despite multiple recent efforts, the majority of the hits discovered by traditional target-based screening showed low efficiency in vivo. Therefore, there is heightened demand for whole-cell based approaches directly using host-pathogen systems. The phenotypic host-pathogen assay described here is based on the monitoring of GFP-expressing Mycobacterium marinum during infection of the amoeba Acanthamoeba castellanii. The assay showed straight-forward medium-throughput scalability, robustness and ease of manipulation, demonstrating its qualities as an efficient compound screening system. Validation with a series of known antitubercular compounds highlighted the advantages of the assay in comparison to previously published macrophage-Mycobacterium tuberculosis-based screening systems. Combination with secondary growth assays based on either GFP-expressing D. discoideum or M. marinum allowed us to further fine-tune compound characterization by distinguishing and quantifying growth inhibition, cytotoxic properties and antibiotic activities of the compounds. The simple and relatively low cost system described here is most suitable to detect anti-infective compounds, whether they present antibiotic activities or not, in which case they might exert anti-virulence or host defense boosting activities, both of which are largely overlooked by classical screening approaches.