We developed a method for quantifying dynamic chemoresponsiveness on the basis of the ventilatory response to pseudorandom binary CO2 stimulation. The dynamic chemoreflex gain (GD) and effective time delay (TDeff) relating breath-to-breath fluctuations in alveolar PCO2 to ventilation were evaluated at frequencies between 0 and 0.05 Hz. Application of the method to simulated "data" showed that estimation errors in GD and TDeff were most likely to be minimized in the range of 0.01-0.03 Hz, corresponding to periodicities of 30-100 s. Estimation of TDeff was generally more susceptible to error than that of GD because of the limited time resolution of the breath-by-breath measurements. In eight awake normal adults, we compared estimates of GD derived from the pseudorandom binary CO2 stimulation test with peripheral and central hypercapnic sensitivities deduced from single-breath and Read rebreathing measurements in the same subject. GD at 0.02 Hz was highly correlated with peripheral hypercapnic sensitivity but poorly correlated with central hypercapnic sensitivity, underscoring the importance of the peripheral chemoreflexes in mediating ventilatory responses to phasic stimuli. Application of the procedure to a different group of 10 healthy volunteers during wakefulness and stage 2 sleep showed decreases in GD in 8 subjects but increases in 2 subjects. However, for the group as a whole, GD and TDeff did not change significantly between wakefulness and sleep. The proposed method may provide information more pertinent to periodic breathing than traditional CO2 response tests do, since the chemoreflex responses to phasic variations in blood gases are likely to be important in determining ventilatory control during sleep.