To evaluate respiratory drive and timing in 11 spontaneously breathing infants anaesthetized with halothane, ventilation was followed before and during CO2 provocation, and occlusion tests were performed. All infants were younger than 6 months of age and their weights ranged from 3.8 to 7.5 kg. All measurements were performed prior to surgery. Tidal volumes (VT) were followed by pneumotachography and end-tidal CO2 concentration [E'CO2) by an in-line capnograph. Occlusion pressure curves were biphasic with an initial fast phase (pressure: P degree fast, duration: T degree fast) followed by a slower phase to the maximal occluded infra-airway pressure (P degree max, T degree max). During CO2 breathing, mean values of P degree fast increased by 75% (P less than 0.001) and of P degree max by 73% (P less than 0.001) compared with at CO2-free breathing. The slope of the fast phase (delta P/delta t) was significantly increased during CO2 breathing while the slow phase was unchanged by the presence of CO2. The P degree fast/P degree max ratio was of the same size before and during CO2 inhalation. Inhalation of CO2 did not influence inspiratory (T1) and expiratory (TE) times during unoccluded breathing. A variable respiratory pattern was revealed during occlusion whilst CO2-free breathing: T degree max was longer than (T1) in nine cases and shorter in two. A more uniform response in ventilatory timing was found at CO2 loaded ventilation and T degree max as well as the total duration of the ventilatory cycle (T degree tot) were significantly longer than (T1) (P less than 0.01) and (Ttot) (P less than 0.05) respectively. The V1/T1 ratio was increased by 66% during CO2 provocation during unoccluded breathing. The net effect of increased inspiratory drive during CO2 breathing resulted in a VT which on average was increased by 67% (P less than 0.001) so that the mean value of E'CO2 only rose by 0.98% (P less than 0.01) from 5.18% before to 6.16% during CO2 breathing. It was concluded that ventilatory compensation to CO2 was adequate, indicating preserved respiratory centre activity. Respiratory timing, however, was unaffected by CO2 indicating a discrepancy between the effects of halothane on respiratory motor centre activity and the bulbopontine pacemaker in these young infants.