Background: This study was design to assess clinical agitation, electroencephalogram (EEG) and autonomic cardiovascular activity changes in children during induction of anesthesia with sevoflurane compared with halothane using noninvasive recording of EEG, heart rate, and finger blood pressure. Methods: Children aged 2-12 yr premedicated with midazolam were randomly assigned to one of three induction techniques: 7% sevoflurane in 100% O-2 (group Sevo(RAPID)); 2%, 4%, 6%, and 7% sevoflurane in 100% O-2 (group Sevo(INCR)); or 1%, 2%, 3%, and 3.5% halothane in 50% N2O-50% O-2 (group Halo(N2O)). An additional group of children who received 7% sevoflurane in 50% N2O-50% O-2 (group Sevo(N2O)) was enrolled after completion of the study. Induction was videotaped EEG, heart rate, and finger blood pressure were continuously recorded daring induction until 5 min after tracheal intubation and analyzed in frequency domain using spectral analysis. Results: Agitation was more frequent when anesthesia was induced with 100% O-2 compared to the mixture of oxygen and nitrous oxide, No seizures were recorded in any group. In the four groups, induction of anesthesia was associated with an increase in EEG total spectral power and a shift toward the low-frequency bands. Sharp slow waves were present on EEG tracings of the three sevoflurane groups, whereas slow waves and fast rhythms (spindles) were observed in the halothane group. Sevoflurane induced a greater withdrawal of parasympathetic activity than halothane and a transient relative increase in sympathetic vascular tone at loss of eyelash reflex. Conclusions: Agitation observed during sevoflurane induction was not associated with seizures. Sevoflurane induction induced a marked inhibition of parasympathetic control of heart rate.