OBJECTIVE Measurement of incidence of neonatal seizures and evaluation of the therapeutic strategy used. METHODS Neonatal intensive care unit, University Hospital, Groningen, the Netherlands. METHODS Retrospective, descriptive. METHODS All patients admitted in a period of 5.5 years who had clinical seizures within 28 days from birth received pharmacological treatment in 4 steps. When a seizure remained clinically uncontrolled, the next step was taken. Step 1: methylphenobarbital 15-35 mg/kg i.v. Step 2: phenytoin 15-35 mg/kg i.v. After an initial loading dose of 15 mg/kg i.v., both the methylphenobarbital and the phenytoin loading dose were stepwise increased until reaching the maximum of 35 mg/kg. Step 3: clonazepam 1-2 x 0,15 mg/kg i.v. Step 4: thiopental 1-2 mg/kg/hour i.v. RESULTS The incidence of neonatal seizures was 121/2068 (5.9%). For assessment of the efficacy of the therapeutic strategy used, data of 37 patients were not available: 16 patients because of missing data, 11 because of deviation from the strategy and 10 because they died during the strategy. Cases of 84 patients could be evaluated for treatment response. The cumulative response rates were 68% (57/84) controlled seizures for methylphenobarbital, 87% (73/84) for phenytoin, 94% (79/84) for clonazepam, and 96% (81/84) for thiopental. For both methylphenobarbital and phenytoin a marked increase in clinical response was seen at increasing loading doses. CONCLUSIONS The stepwise therapeutic strategy used in this study proved to be successful in most cases of neonatal seizures. The anticonvulsive action of both methylphenobarbital and phenytoin appeared to be dose-related.