When researchers screen potential tuberculosis drugs, they often end up with too many options. Some look promising but later prove to be costly dead ends.