Modulation of heart rate (HR) during transient hyperoxia, hypoxia, and hypercapnia was studied in 46 healthy term infants on 103 occasions (postnatal d 2 to 82). Twenty-three infants had smoking mothers (median, 11 cigarettes/d). Transient chemoreceptor stimuli (100% O(2), 15% O(2), or 3% CO(2)) were presented repeatedly during quiet sleep. Beat-by-beat HR and breath-by-breath ventilation were recorded continuously. The coherently averaged HR and ventilation responses to each stimulus were calculated for each infant at each age. Outcome variables (HR change from baseline to end of stimulation, maximum HR change, and time to half-maximum) were analyzed by ANOVA. Overall, HR declined during hyperoxia (median change, 4.2 beats/min) and rose during hypoxia (median change, 4.2 beats/min) and hypercapnia (median change, 4.6 beats/min). The percentage change in HR was positively correlated with the percentage change in ventilation (p < 0.001). Increasing number of cigarettes smoked by the mother was correlated with deeper HR declines and smaller HR rises (p = 0.02). For the population as a whole, the HR response lagged 3.8 s behind the ventilatory response during hyperoxia and hypoxia (p < 0.001), whereas during hypercapnia there was no significant lag. The lag in HR response in the smoke-exposed group was 2.5 s greater than that in the control group for all three stimuli (p = 0.001), and the difference increased with the number of cigarettes smoked by the mother (p < 0.01). Both pulmonary reflexes and the type of the chemoreceptor stimulus seemed to influence HR. Maternal smoking affected the magnitude and time-course of the HR response in a dose-dependent manner.