Twenty three children with recurrent supraventricular tachycardia were treated with flecainide. Twenty one of these received intravenous treatment during an attack (2 mg/kg over 10 minutes). The tachycardia was terminated in 17. After an intravenous bolus of flecainide, blood samples were drawn at regular intervals for analysis of flecainide concentration over 48 hours. Pharmacokinetic variables were calculated--median terminal half life 7.5 hours, median volume of distribution 6.2 l/kg, and median plasma clearance 7.2 ml/min/kg. There was a significant correlation between half life and age. Twenty of the children received long term treatment with an oral preparation of flecainide to prevent further attacks. Twelve had no further attacks and 16 were considered to have good control. Two children suffered potentially serious arrhythmogenic effects soon after the start of oral treatment and flecainide had to be stopped. During oral treatment regular blood samples were drawn and plasma concentrations were analysed to assess the therapeutic range. This did not differ substantially from that proposed in adults (400-800 micrograms/l). Eight children were electively withdrawn from oral flecainide to see whether they really needed it. Blood samples for measurement of flecainide concentration were drawn after their last oral dose. Pharmacokinetic variables were calculated: time to maximum concentration 2 hours, median terminal half life 7.9 hours. For the combined data from patients receiving intravenous and oral treatment there was a significant correlation between elimination half life and age. An intravenous dose of 2 mg/kg over at least 10 minutes and an initial oral dose of 6 mg/kg/day in three divided doses is recommended. Treatment should be started in hospital so that children in whom the drug may be arrhythmogenic can be identified and plasma concentrations measured to identify patients in whom lack of efficacy is caused by underdosage.