Hypermetabolism of phenytoin is not frequently recognized as a cause of treatment failure. We report the case of a 37-year-old male in whom detailed pharmacokinetic investigation revealed that hypermetabolism, rather than lack of compliance or poor absorption, was responsible for low plasma levels of phenytoin. An increase of his daily dose of phenytoin to 800 mg resulted in adequate plasma levels and good seizure control. Additional studies with two model drugs metabolized by the liver--aminopyrine and antipyrine--showed that he was also a fast metabolizer for these substrates, suggesting a nonspecific induction of hepatic drug metabolizing enzymes. Low plasma phenytoin levels should not be systematically ascribed to lack of compliance, and increased phenytoin metabolism should be considered as an occasional cause of treatment failure.