Hypnotic endpoints and/or EEG variables, e.g. bispectral index, 95% spectral edge frequency and median frequency, have been studied to monitor anaesthetic (hypnotic) depth during total intravenous anaesthesia. In this study, the relation between the hypnotic endpoints of unresponsiveness to verbal commands, loss of eyelash reflex and body movement response to mechanical nasal membrane stimulation vs. bispectral index, 95% spectral edge frequency and median frequency during propofol anaesthesia with or without fentanyl is presented. Forty-two patients were randomly assigned to receive either propofol infusion, 30 mg kg−1 h−1 (n=22), or propofol infusion, 30 mg kg−1 h−1+fentanyl bolus, 2 μg kg−1 i.v. (n=20). Bispectral index, 95% spectral edge frequency and median frequency and propofol doses were monitored and recorded at unresponsiveness to verbal commands, loss of eyelash reflex and inhibition of nasal body movement response. The bispectral index values were significantly higher in the propofol+fentanyl compared with the propofol group, i.e. 74.7±10.9, 73.1±10.5 and 47.1±9.2 vs. 65.8±9.8, 59.6±10 and 33.8±5.7 at unresponsiveness to verbal commands, loss of eyelash reflex and inhibition of nasal body movement response respectively. Doses of propofol for achieving the hypnotic endpoints were significantly lower in the propofol+fentanyl compared with the propofol group. Plasma propofol concentrations at inhibition of nasal body movement response were lower in the propofol+fentanyl compared with the propofol group (9.2±2.0 μg mL−1 vs. 14.1± 4.2 μg mL−1). Our results suggest that fentanyl pretreatment potentiates the effects of propofol and achieves the hypnotic endpoints at higher bispectral index values and lower propofol doses and concentrations (measured at inhibition of nasal body movement response).