Leflunomide is a disease-modifying antirheumatic drug (DMARD) for the treatment of rheumatoid arthritis (RA). Structurally, it is a derivative of 5-methylisoxazole-4-carboxamide. Upon metabolism, the N-O bond in the isoxazole ring is cleaved to form the active metabolite, teriflunomide, which was recently approved by the FDA for the treatment of multiple sclerosis. Both leflunomide and teriflunomide inhibit dihydroorotate dehydrogenase (DHODH) thereby inhibiingt the synthesis of pyrimidine. For both drugs, the two major concerns are potential liver toxicity and teratogenicity. It was suspected that these undesirable effects might be related to the cleavage of the N-O bond. We herein summarize the metabolites-toxicity issues related to leflunomide/teriflunomide and discuss two related molecular platforms, UTL-4 and UTL-5. UTL-4 compounds are based on the same scaffold of leflunomide; their toxicological and pharmacological effects are not significantly different from those of leflunomide/teriflunomide. In UTL-5 series, the leflunomide scaffold is changed into 5-methylisoxazole-3-carboxamide. Unlike leflunomide, the N-O bond of a UTL-5 compound, UTL-5b, is not cleaved upon metabolism; instead, the peptide bond is cleaved to form its major metabolites. UTL-5b and its metabolites do not inhibit DHODH in vitro. In addition, UTL-5b and all other UTL-5 compounds have lower acute toxicity than leflunomide/teriflunomide. Furthermore, from leflunomide to UTL-5b/UTL-5g, the potential liver toxicity becomes liver protective effect. With the reduced toxicity, UTL-5 compounds still maintain significant pharmacological effects including anti-inflammatory and antiarthritic effects. In summary, our observations provide a valuable direction in drug optimization based on the modification of the leflunomide scaffold.