Ventilatory responses to CO2 ($$\Delta \dot V_I /\Delta P_{CO_2 } $$) were measured half, one, two, and four hours after infusions of thiopentone, morphine, fentanyl and saline in healthy men in order to test the idea that variation in clinical recovery and control of breathing after anaesthetic drugs are associated with interindividual differences in control measurements of$$\Delta \dot V_I /\Delta P_{CO_2 } $$. Ventilatory response to CO2 was profoundly reduced one half hour after each drug, in contrast to the observation during air breathing that ventilation and end tidal PCO2 had returned to within 10 per cent of control. Mean$$\Delta \dot V_I /\Delta P_{CO_2 } $$ increased progressively at one, two, and four hours, returning to near control after thiopentone, but remaining less than 80 per cent of control four hours after morphine and fentanyl. From the regression equations of each ventilatory response, ventilation at Pco2 f 58 and 70 mmHg ($$\dot V_{I_{58} } $$ and$$\dot V_{I_{70} } $$) were computed to estimate displacement of the response curves by the drugs. Following thiopentone there was no significant change of$$\dot V_{I_{58} } $$. In contrast there was a highly significant fall of$$\dot V_{I_{58} } $$ one half hour after fentanyl (p < 0.01), with progressive return towards control at one, two, and four hours; similar changes were observed after morphine. For each drug, changes of$$\dot V_{I_{70} } $$ were substantially greater than corresponding changes of$$\dot V_{I_{58} } $$. At all times during these recovery measurements, subjects were conscious and co-operative and, by traditional clinical criteria, were judged to have recovered from the effects of the drugs. Differences between high and low responding subjects were assessed by plotting control measurements against values obtained half and one hour after drugs. No systematic differences were found. These findings suggest that$$\Delta \dot V_I /\Delta P_{CO_2 } $$ is a sensitive indicator of central nervous activity, but do not support the concepts that individuals with low$$\Delta \dot V_I /\Delta P_{CO_2 } $$ might be more susceptible to the ventilatory depressant effects of anaesthetic drugs, or that low$$\Delta \dot V_I /\Delta P_{CO_2 } $$ might be associated with delayed return of spontaneous breathing after general anaesthesia. Plasma thiopentone levels at half, one, and four hours were highly reproducible, in contrast to the wide variation of$$\Delta \dot V_I /\Delta P_{CO_2 } $$ among subjects in this study. These findings together support the notion that wide variation in clinical recovery from anaesthesia may have a primary physiological basis in addition to variation caused by interindividual differences in drug dosage, biotransformation and excretion.