SUMMARY1. The respiratory responses to C02 inhalation were measured in New Zealand White rabbits. Prior to testing, the rabbits either received drinking water ad libitum (C), or were subjected to 72 hr water deprivation (WD). The rabbits were tested in a climatic chamber at either 20 0, (C20, WD20) or 30 0C (C30, WD30).2. C02 exposure caused increases in both tidal volume (VT) and minute volume (VE). The direction and magnitude of changes in respiratory frequency (f), however, were dependent upon treatment.3. Linear regressions were drawn for f vs. % C02 in inhaled gas. Treatment C20 showed zero slope, treatment C30 a large negative slope, and treatments WD20 and WD30 a significant positive slope.4. The slopes ofthe regression lines obtained from plottingf vs. % C02 were plotted against their intercepts for each rabbit under each treatment condition. This yielded a highly significant linear regression from which it could be concluded that the frequency response to C02 exposure is dependent upon the initial frequency when the rabbits are breathing air.5. It is suggested that frequency is shifted during hypercapnia towards an optimal value which is dependent upon the existing % C02 in the inspired gas.6. Comparison of the results obtained for the rabbit with results reported for the conscious dog and sheep suggest that a similar relationship may hold for these species.