Ｏｂｊｅｃｔｉｖｅ　To evaluate anesthesia effect of three different sevoflurane inhalational induction methods in children.　Methods　Ninty children patients undergoing elective tonsillectomy or adenoidectomy (ASA Ⅰ) were randomly allocated into incremental inhalation group (group A), fixed-concentration inhalation group (group B) and multiple-deep-breath inhalation group(group C)(n=30). MAP, HR and SpO2 at the moments of patients entering into the anaesthetic induction room (T0), lose of eyelash refex(T1) and intravnous cannulation after 160 s of anaesthetic inhalation (T2) were recorded. Intravenous cannulation was performed if movement and eyelash reflex did not occur after 160 s of anaesthetic inhalation. Induction cooperation, induction time, face mask application to loss of the eyelash reflex, vital signs(MAP, HR, SpO2), induction complications, recovery time(Tc), extubation time(Td), recovery agitation, recovery complications were recorded.　Results　The time of loss consciousness and eyelash reflex in group A(58.1±0.3) s was the longest but the shortest in group C(33.8±0.8) s (P<0.05). Patient movement during cannulation in group A(96.7%) was higher than that in group B(16.7%) and C(13.3%) (P<0.05). Patient movement during induction in group A(100%) was higher than that in group B(13.3%) and C(10%) (P<0.05). Bradycardia and salivation only happened in one child from group A. Induction compliance checklist in group B(0.20±0.07) was the lowest, whereas the highest in group C(1.43±0.13)(P<0.05).　Conclusions　We recommended the fixed-concentration of sevoflurane (8%) for inhalational induction of anesthesia in children.