Clostridium difficile infection (CDI) is the leading cause of healthcare-associated infection in the United States. Therefore, development of novel treatments for CDI is a high priority. Toward this goal, we began in vitro screening of a structurally diverse in-house library of 67 compounds against two pathogenic C. difficile strains (ATCC BAA 1870 and ATCC 43255), which yielded a hit compound, 2-methyl-8-nitroquinazolin-4(3H)-one (2) with moderate potency (MIC = 312/156 μM). Optimization of 2 gave lead compound 6a (2-methyl-7-nitrothieno[3,2-d]pyrimidin-4(3H)-one) with improved potency (MIC = 19/38 μM), selectivity over normal gut microflora, CC50s > 606 μM against mammalian cell lines, and acceptable stability in simulated gastric and intestinal fluid. Further optimization of 6a at C2-, N3-, C4-, and C7-positions resulted in a library of >50 compounds with MICs ranging from 3 to 800 μM against clinical isolates of C. difficile. Compound 8f (MIC = 3/6 μM) was identified as a promising lead for further optimization.