OBJECTIVE To describe 2 cases of clinically significant phenytoin removal during continuous venovenous hemofiltration (CVVH) and review the relevant literature regarding phenytoin removal by renal replacement modalities. METHODS A 64-year-old female with chronic kidney disease and cirrhosis was admitted to the intensive care unit (ICU) with a traumatic subdural hematoma and seizures. The patient received a loading dose of intravenous phenytoin 1000 mg, followed by maintenance intravenous administration of phenytoin 100 mg and levetiracetam 250 mg every 12 hours. CVVH was initiated for acidosis. A 63-year-old male was admitted to the ICU after cardiac surgery complicated by hypotension. CVVH was initiated for fluid overload, and phenytoin was initiated 3 days later for seizures. A loading dose of intravenous phenytoin 2700 mg was administered, followed by maintenance dosing of intravenous phenytoin 150 mg every 8 hours. Concentrations of unbound phenytoin in serum and CVVH effluent samples were measured during concomitant treatment in each patient. In both patients, serum and effluent concentrations of unbound phenytoin fell steadily while they were on CVVH. Clearance of phenytoin by CVVH was calculated, as was the daily removal of phenytoin, as a percentage of total daily phenytoin dosage during each sampling period. Phenytoin clearance by CVVH ranged from 11 to 13 mL/min in these patients. CONCLUSIONS The clearance of phenytoin with CVVH in these 2 patients was much higher than the renal clearance of phenytoin reported in healthy volunteers with normal renal function. Previous case reports have demonstrated that only small, clinically insignificant amounts of phenytoin are removed by hemodialysis, and the only published report of phenytoin removal by continuous renal replacement therapy used hemofiltration rates much lower than those used in the 2 cases described here. CONCLUSIONS These cases demonstrate that a substantial amount-approximately 30%-of total daily phenytoin dose may be removed by CVVH, and patients may require higher than expected empiric doses. Phenytoin concentrations should be closely monitored in critically ill patients receiving CVVH.