A decrease in volatile anesthetic consumption has been demonstrated using bispectral index (BIS), whereas data concerning spectral entropy are lacking. One hundred and forty adult patients scheduled for surgical procedures lasting more than 1 h were prospectively randomized to receive an anesthetic controlled either by BIS or by spectral entropy or solely by clinical variables. Anesthesia was induced with propofol and sufentanil. Sufentanil was infused continuously thereafter. Sevoflurane was administered in 1 L/min O2/N2O. The sevoflurane concentration was adjusted according to conventional clinical variables in the standard practice group, whereas the 40-60 interval was applied for the BIS and spectral entropy-guided groups. The sevoflurane vaporizer was weighed before and after anesthesia, and consumption was calculated. Groups were comparable for demographic data except for weight (heavier in the spectral entropy-guided group, P < 0.05). Compared with standard practice, patients with BIS or spectral entropy monitoring required 29% less sevoflurane (normalized sevoflurane consumption to the weights of the patients and to the durations of anesthesia; both P < 0.03) and a similar sufentanil dose. An unintended improvement in the standard practice group (positive bias) was observed. In conclusion, BIS and spectral entropy monitoring have the same sparing effect of sevoflurane.