The activity of sympathetic premotor neurons in the rostral ventrolateral medulla (subretrofacial nucleus) supports sympathetic vasomotor tone, but the factors that drive these premotor neurons' activity have not been determined. This study examines whether either direct interconnections between subretrofacial neurons or synchronizing common inputs to them are important for generating their tonic activity. Simultaneous extracellular single-unit recordings were made from 32 pairs of sympathetic premotor neurons in the subretrofacial nucleus of chloralose-anesthetized cats. Paired spike trains were either separated by spike shape from a single-electrode recording (14 pairs) or recorded from two electrodes less than 250 microm apart (18 pairs). All neurons were inhibited by carotid baroreceptor stimulation and most had a spinal axon proven by antidromic stimulation from the spinal cord. Autocorrelation, inter-spike interval, and cardiac cycle-triggered histograms were constructed from the spontaneous activity of each neuron, and cross-correlation histograms covering several time scales were generated for each neuron pair. No significant peaks or troughs were found in short-term cross-correlation histograms (2 ms bins, +/-100 ms range), providing no support for important local synaptic interactions. On an intermediate time scale (20 ms bins, +/-1 s range), cross-correlation revealed two patterns indicating shared, synchronizing inputs. Repeating peaks and troughs (19/32 pairs) were due to the two neurons' common cardiac rhythmicity, of presumed baroreceptor origin. Single, zero time-spanning peaks of 40--180 ms width were seen in 5/32 cases. Calculations based on the prevalence and strength of these synchronizing inputs indicate that most of the ensemble spike activity of the subretrofacial neuron population is derived from asynchronous sources (be they intrinsic or extrinsic). If synchronizing sources such as neuronal oscillators were responsible for more than a minor part of the drive, they would be multiple, dispersed, and weak.