Tuberculosis (TB), caused by Mycobacterium tuberculosis, is the biggest infectious killer worldwide, with 9.6 million new infections and 1.3 million deaths per year. M. tuberculosis can exist in a latent state in the lung within hypoxic, necrotic granulomas. These granulomas present a major hurdle for both immunological clearance and drug treatment. As such, many of the drugs currently available to treat TB have limited efficacy due to their inability to reach the center of granulomas where most bacteria reside. Advances in TB drug discovery are slow, which may be due to ineffective models in novel TB compounds may be identified. Granulomas share their hypoxic, necrotic characteristics with avascular tumors, which also face drug penetration barriers. Recent work has used a three-dimensional in vitro spheroid model that is hypoxic and necrotic at the core to model assess the ability of novel anti-cancer drugs to reach the center of tumors where the target cells reside. The current study aimed to adapt this model to create a novel in vitro infection model that could be used to assess the efficacy of novel anti-TB compounds that more accurately reflects the conditions of a granuloma in vitro than traditional drug-screening methods. Spheroids were formed using DLD-1 cells infected with fluorescent M. tuberculosis H37Ra; confocal and two-photon microscopy revealed three-dimensional aggregates of cells that maintained their integrity when infected with bacteria, which was distributed evenly throughout the structures. Sectioning spheroids made with hypoxia-reporter cell lines revealed that infected spheroids appear hypoxic and necrotic at their core. Both novel and standard anti-TB drugs were assessed in this model, and a reduced bacterial load was observed by flow cytometry and enumeration of bacteria on agar. This study identifies a novel in vitro three-dimensional infection model that can be used to assess novel drug efficacy that better reflects the in vivo conditions of TB infection.