Use of the benzodiazepine antagonist flumazenil may inhibit the effects of benzodiazepines in a competitive manner. The only known partially agonistic effect of flumazenil is a weak anticonvulsive action at high doses. However, reports have claimed that flumazenil reduces the MAC of isoflurane in animal studies. Other reports have found that antagonizing midazolam-induced sedation or anesthesia by flumazenil led to an increase in respiratory depression. The aim of this study was to examine whether flumazenil i.v. increases fentanyl-induced respiratory depression. METHODS. In two separate sessions, ten healthy young volunteers were given either 0.0027 mg/kg fentanyl alone or 0.0027 mg/kg and 1 mg flumazenil i.v. over 4 min each time. The CO2 rebreathing method was used to determine the ventilatory response. RESULTS. Fentanyl alone brought about a significant reduction in CO2 response, characterized by a shift to the right and a decrease in the slope of the rebreathing curve (from 1.95 +/- 0.76 l.min-1.mmHg-1 to 0.86 +/- 0.53 l.min-1.mmHg-1). The infusion of additional flumazenil caused similarly significant respiratory depression (from 2.21 +/- 1.0 l.min-1.mmHg-1 to 0.77 +/- 0.38 l.min-1.mmHg-1). In both groups changes persisted for at least 120 min. No statistically significant differences between the two groups could be detected. CONCLUSION. Flumazenil does not enhance fentanyl-induced respiratory depression. Flumazenil's weak, partially agonistic action is therefore of no clinical importance.