OBJECTIVE To determine the incidence of abstinence syndrome in children interned in the Pediatric Intensive Care Unit (PICU) in fentanyl use and midazolam. METHODS Evaluation of 36 children interned in PICU of the Hospital São Paulo--Federal University of São Paulo, in the period from March to September 1997, with age varying from 5 days to 22 months (22 masc: 14 fem) who used fentanyl use and midazolam for more than 24 hours. Used the Escore Neonatal of Abstinence adapted by Finnegan determines the occurrence of abstinence syndrome in was used to children 2 years old or less. Sustain larger or equal for 8 is considered as abstinence syndrome. Correlated the abstinence syndrome with the accumulated total dose, infusion velocity, daily dose and time of use of the fentanyl and midazolam. RESULTS Certain abstinence syndrome in 18 (50%) of the 36 children. Applied Mann Whitney's statistical test to compare the groups with and without abstinence syndrome. Dose accumulated of fentanyl total (5732.7 +/- 5114.91 vs 624.2 +/- 591.2 mcg, p < 0.005), dose daily of fentanyl (98.54 +/- 6.12 vs 36.23 +/- 23.42 mcg/Kg/dia, p < 0.005), velocity of infusion of the fentayl (4.09 +/- 2.75 vs 1.5 +/- 0.95 mcg/Kg/hora, p < 0.005), time of use of the fentanyl (10.2 +/- 5.1 vs 3.16 +/- 1.09 days, p < 0.005), dose accumulated of midazolam total (118.8 +/- 86.97 vs 20.03 +/- 14.79 mg, p < 0.005), dose daily of midazolam (2.32 +/- 0.86 vs 1.21 +/- 0.68 mg/Kg/dia, p < 0.005), velocity of midazolam infusion (0.13 +/- 0.16 vs 0.05 +/- 0.02 mg/Kg/hora, p < 0.005) and time of use of the midazolam (9.20 +/- 4.67 vs 2.55 +/- 1.54 days, p < 0.005) they were considered significant. CONCLUSIONS The abstinence syndrome presents an elevated incidence in children interned in PICU owing to the interruption of the fentanyl infusion and midazolam and in these patients it was related with the dose and the time of use.