In 49 patients undergoing open-heart surgery we compared the electroencephalographic (EEG) effects of high-dose morphine, fentanyl, or sufentanil with O2, using two computerized analysis and display techniques: a period analysis (the Klein method) and an aperiodic analysis (the Neurometrics monitor). During fentanyl or sufentanil anesthesia, both techniques revealed a general decrease in frequency, shown by the aperiodic analysis primarily as a marked increase in the very low frequency range: an increase in the 1-Hz bin (TP1, in μv2) from 2.80 × 104 ± 3.20 × 104 (SD) to 45.1 × 104 ± 27.2 × 104 for fentanyl and from 3.11 × 104 ± 2.83 × 104 to 52.8 × 104 for sufentanil. The cumulative percent power at 3 Hz (CP3) increased from 27.2 ± 6.8 to 83.0 ± 11.0 for fentanyl and from 22.7 ± 5.2 to 85.1 ± 10.4 for sufentanil, while the frequency at 90% cumulative percent power (F90, in Hz) decreased from 17.8 ± 2.9 to 7.9 ± 2.8 for fentanyl and 16.4 ± 5.2 to 5.6 ± 4.3 for sufentanil. The changes with morphine were less obvious, with some attenuation of high-frequency power shown by the Klein method, and an increase from 24.1 ± 8.6 to 59.3 ± 20.7 with CP3, but no change in TP1. Low-frequency power with the period analysis and TP1 with the aperiodic analysis decreased between laryngoscopy and the incisions with fentanyl and sufentanil. We conclude that these narcotics given in high doses markedly alter the EEG, that the changes with fentanyl and sufentanil are greater than those with morphine, that computerized analysis and display techniques can quantitate these changes, that--- using the EEG of other general anesthetics as a guide---the patients were adequately anesthetized, and that, given the appropriate variables, the EEG probably reflects the depth of anesthesia with high-dose narcotics.