The authors determined the effect of intravenous lidocaine, both as a bolus and as an infusion, on the ventilatory response to CO2. Bolus injection of 1.5 mg/kg lidocaine caused a decrease in the slope of the CO2 ventilatory response curve from 2.66 +/- 0.30 (mean +/- SEM) to 1.31 +/- 0.44 1 X min-1 X mmHg-1 within 90 s; the effect was transient, with slope returning to 2.39 +/- 0.83 1 X min-1 X mmHg-1 150 s after injection. The transient, subconvulsive lidocaine concentrations present during ventilatory depression (8.9 +/- 2.0 micrograms/ml) may be sufficient to desensitize the medullary ventilatory control centers. Lidocaine infusion at the rate of 60 micrograms X kg-1 X min-1 (serum lidocaine concentrations of 3.5 +/- 0.2 micrograms/ml) increased the slope of steady state CO2 response curves from 2.89 +/- 0.29 to 4.17 +/- 0.44 1 X min-1 X mmHg-1 (P less than 0.05); with discontinuation of the infusion, slope returned to 3.18 +/- 0.33 1 X min-1 X mmHg-1 (P less than 0.05). The authors conclude that bolus injection of lidocaine transiently can depress ventilatory control, however, rapid redistribution of lidocaine makes this a transient phenomenon that can be treated with supplemental oxygen if necessary. The increased CO2 sensitivity observed during lidocaine infusion suggests that studies of ventilatory control in patients receiving conduction anesthetics must take into account the direct effect of absorbed anesthetics on ventilatory control.