The hypothesis that calcium channel blockers can potentiate and prolong the anti-nociceptive effects of opioids was tested. Forty-five men scheduled for aorto-coronary bypass operation received fentanyl according to their individual demands (haemodynamics, clinical parameters). The patients were allocated at random into three groups receiving either nimodipine 1.0 microgram kg-1 min-1 (Group 1, n = 15), nifedipine 0.70 microgram kg-1 min-1 (Group 2, n = 15), or no calcium channel blocker (Group 3, n = 15). Cerebral activity was monitored using a computerized spectral analysing system before and during the operation. The total amount of fentanyl required was significantly lower in the nimodipine group than the control group (-71%, P less than 0.001), whereas the nifedipine group did not differ from the control group. Quality of intra-operative anaesthesia was comparable in the three groups with respect to clinical observations (amnesia, sweat, tears, pupils), and the post-operative course was similar in all patients as well. Cerebral activity during the nimodipine-supplemented opioid anaesthesia was higher in the faster frequency bands (13-30 Hz). Power level in the beta range was most pronounced in Group 1, whereas power in the alpha range was similar in the calcium channel-blocker groups. The major conclusion was that nimodipine but not nifedipine administration can reduce fentanyl requirements during surgical procedures without influencing the quality of anaesthesia.