13Mycobacterium tuberculosis possesses a large number of genes of unknown or merely 14 predicted function, undermining fundamental understanding of pathogenicity and drug 15 susceptibility. To address this challenge, we developed a high-throughput functional 16 genomics approach combining inducible CRISPR-interference and image-based analyses of 17 morphological features and sub-cellular molecular localizations in the related non-pathogen, 18M. smegmatis. Applying automated imaging and analysis to an arrayed library of 272 19 essential gene knockdown mutants, we derive robust, quantitative descriptions of bacillary 20 morphologies consequent on gene silencing. Leveraging statistical-learning, we demonstrate 21 that functionally related genes cluster by morphotypic similarity and that this information 22can be used to infer gene function. Exploiting this observation, we reveal a previously 23 unknown restriction-modification system, and identify filamentation as a defining 24 mycobacterial response to histidine starvation. Our results support the application of large-25 scale image-based analyses for mycobacterial functional genomics, simultaneously 26 establishing the utility of this approach for drug mechanism-of-action studies.