In cats anesthetized with alpha-chloralose and urethane, the medulla oblongata was explored with electrical and/or chemical stimulation for vasopressor actions. Chemical stimulation included semimicroinjection of sodium glutamate or microinjection of dl-homocysteic acid (DLH). The dorsal (DM), particularly its dorsomedial (DMM) portion, and the ventrolateral (VLM) regions were found to be most sensitive to both electrical and chemical stimulation. In between these two regions there was an area in the ventral portion of the gigantocellular and the parvocellular reticular nuclei that was sensitive to electrical and somewhat sensitive also to chemical stimulation. The responses to chemical stimulation indicate the existence of perikarya in these three areas for vasopressor actions. The most active response of VLM followed microinjection of DLH into the region of nu. paragigantocellularis lateralis (PGL) and subretrofacial nu. These responses were about 30% greater than those from DM where the most active area was found in the DMM region including the nu. and tractus solitarius. When an extensive area of VLM was lesioned bilaterally with kainic acid (KA, 1 μg/100 nl), the resting systemic arterial blood pressure (SAP) fell 42% and the pressor response on DM stimulation fell by more than 80%. In contrast, after similar KA bilateral destruction of DM the resting SAP fell 38% and the pressor responses from VLM decreased 67%. When only a limited area in the PGL/subretrofacial nu. of the rostral VLM (3-4 pressor sites, 1 mm apart) on one side was lesioned, the resting SAP remained unaltered while the pressor response on the ipsilateral DMM decreased 76%. On the other hand, when a similar smaller KA lesion (3-4 sites) was made in the DMM, although the resting SAP did not change, the pressor response on stimulation of the ipsilateral rostral VLM decreased 28%. Further lesioning of the contralateral DMM then decreased the rostral VLM pressor response 62% without much alternation in the resting SAP. Results of the present experiments are in accordance with our previous results that neuronal perikarya for vasopressor action exist not only in VLM but also in DM (18,34) with the function of VLM slightly predominant over DM. It seems highly possible that reciprocal innervation exists between DM and VLM, at least between DMM and the PGL/subretrofacial nu. region.