The respiratory control systems of the predominantly lung-breathing turtle. Pseudemys scripta, and the bimodal breathing frog, Rana catesbeiana, are compared. At constant temperature, control is effected in both species primarily by pulmonary ventilation, and appears to be similar in basic organization to the mammalian system. The skin CO2 exchange of the frog is largely passive and apparently under little adaptive control. CO2 exchange at different temperature is analyzed by means of the conductance equations: convective conductance [MCO2 = (VA/RT).pACO2] for lung exchange, and diffusive conductance (MCO2 = GCO2.transcutaneous deltaPCO2) for skin exchange. Predictions of conductance values, based on data for PaCO2 and MO2, agree well with measured values of both the turtle's ventilation and the frog's skin CO2 conductance at different temperatures. In each case the conductances increase only moderately, if at all, with temperature, a result that can be attributed to the similar Q10 values of PaCO2 and MO2. Because the conductances exhibit this blunted temperature dependence, both lung and skin breathers are provided with the respiratory effector mechanisms necessary to regulate PaCO2 and pHa at their characteristic values over the ectothermic temperature range.