A series of N-(phenylacetyl)trifluoromethanesulfonamides (3a-g) was prepared according to the Topliss scheme in order to determine if aryl substituents would influence anticonvulsant activity. In initial (phase I) screening and quantitative (phase II) evaluation, all seven compounds exhibited significant activity against MES- and scMet-induced seizures. N-(Phenylacetyl)trifluoromethanesulfonamide (3a) was then advanced through five additional testing phases (phases III-VII). Compound 3a displayed good oral bioavailability, low toxicity, and a larger protective index in mice than the prototype drugs, phenytoin, phenobarbital, valproate, and ethosuximide. Additionally, 3a exhibited a longer time to peak effect in all tests and a greater 24-h margin of safety (HD(50)/ED(50)) than the prototypes. Compound 3a blocked picrotoxin-induced seizures but was ineffective against seizures induced by bicuculline or strychnine. In vitro receptor binding studies revealed that 3a did not displace [(3)H]-labeled gamma-aminobutyric acid or [(3)H]-labeled flunitrazepam, and tolerance did not develop during a 5-day chronic administration.