In vitro brainstem spinal cord preparations isolated from newborn rats were used to separately test the effects of modifications of FCO2 and pH of artificial cerebrospinal fluid on the frequency and amplitude of spinal respiratory activity recorded from C2-C8 ventral roots. Different substances such as L-glutamic acid (3 x 10[-3] M), N-methyl-D-aspartic acid (5 x 5 x 10[-6] M), amphetamine (6 mg/100 ml), 5-hydroxytryptophane (10[-3] M), or modified K+ (10[-3] M) were tested for their capacity to elicit stable changes in spinal respiratory activity over a long time period (more than 30 min) and with high frequency of occurrence, i.e., in at least 50% of the cases. None of the above drugs were found to be suitable for the investigation of the chemosensitivity of the spinal respiratory generator (sRG) because they were only able to maintain spinal respiratory activity for around 15 min. Given these data, the previously used procedure of activation through initial deep diethyl ether anaesthesia of newborn rats was employed  to test the chemosensitivity of the sRG because this treatment resulted in the maintenance of spinal respiratory activity with a regular pattern for 30 min, even if it occurred in only 25% of the preparations. After an increase in FCO2 from 5 to 7% (at constant pH 7.4), a significant (p < 0.05) enhancement of the mean frequency was observed on spinal respiratory bursting in both brainstem spinal cord and isolated spinal cord preparations. The changes in burst amplitude, however, were quite variable from one experiment to the other. At constant FCO2 (5%), a decrease in pH from 7.4 to 7.2 enhanced spinal respiratory frequency on brainstem spinal cord or isolated spinal cord preparations, while an increase in pH from 7.4 to 7.6 decreased it. Under these pH conditions, we did not observe any reproducible variations in spinal burst amplitude. From these results, we conclude that this spinal generator is chemosensitive to both CO2 and [H+], suggesting that it belongs to the respiratory system. Our data provide evidence for the existence of spinal CO2 and/or H+ chemoreceptors.