In an attempt to identify new inhibitors of the growth of Mycobacterium tuberculosis (MTB), the causative agent of tuberculosis, a procedure for the generation, design, and screening of a ligand-based virtual library was applied. This used both an in silico protocol centered on a recursive partitioning (RP) model described herein, and a pharmacophoric model for antitubercular agents previously generated by our research group. Two candidates emerged from databases of commercially available compounds, both characterized by a minimum inhibitory concentration (MIC) of 25 microg mL(-1). Based on these compounds, two series of derivatives were synthesized by both parallel solution-phase and microwave-assisted synthesis, leading to enhanced antimycobacterial activity. During both the design and synthesis, attention was focused on the efficient allocation of available resources with the aim of reducing the overall costs associated with calculation and synthesis.