Weconducted a larg e-scale retrosp eetive study to comp are the surgical ejJieaey, praetieal utility, saf ety, and eost-eJJeetiveness ofultrasonic harm onic sealpeItonsillectomy, hot electrocautery, and coldsurg ical dissection. We based our jindings on the length of operating time, comp lication rates, the length of hospital stay f or patients with comp lications, and relative costs. We then cotnp ared our jindings with thos e published in earlier reports, non e of which were based on a three-way comp ariso n. Our study population was made up of316 patients-175 ma les and J4J fetnales aged J to 23 years (mean : 7.3)-who had undergone adenotonsillectomy OI' tonsillectomy alone at our tertiary eare ch ildren s hosp ital betw een Sept. J, 2000, and Aug. 3J, 200J. The harmonic sealp ei was used on 75 patients (23. 7%), electrocautery on J09 patients (34.5%), and cold surgical disseetion onJ32 (4 J.8%). The mean length ofop erating timefor adenotonsillectomy was 42.4 (n = 70), 43.0 (n = J03), and 49.2 (n = 95) minutes, respeetively ; the eorresp onding times f or tonsillectomy alon e were 23.6 (n = 5), 30.2 (n = 6), and 35.3 (n = 37) minutes. Overa/! comp lication rates were 2.7, 5.5, and 6.J%, respe etively. Hospital stays fo r imm ediate « 24 hr) postoperative bleedin g averaged 2.0, J.O, and 0.7 days, resp ectively, and stays f or dehydration averaged 1.