This is the first of a series of four papers that describes the use of a sensitive "microculture" procedure for examining the neurotransmitter profile of a neuron by assaying the transmitter(s) it releases. Sympathetic principal neurons isolated from the superior cervical ganglia of neonatal or adult rats were grown for 10 d to several months on small islands of cardiac myocytes (island diameter, ca. 0.5 mm). To assay transmitter status a neuron and a myocyte in the same microculture were impaled with microelectrodes, the neuron was stimulated and the pharmacology of the effect(s) on the group of electrically coupled myocytes, and on the neuron itself, was investigated. Because the growing axonal processes were confined to the island, the innervation of the myocytes became dense; transmission from neuron to myocytes occurred reliably and was often intense. Most experiments were done on islands containing only a single neuron so that the observed effect(s) on the myocytes could be confidently assigned to that neuron. After the physiological assay, the fine structure or cytochemistry of the neuron was often examined. With single-neuron microcultures the physiology and anatomy of the neuron, including the fine structure of its synaptic endings and varicosities, could be correlated unambiguously. During the course of this work, we have observed five pharmacologically distinct effects exerted on the myocytes by either neonate- or adult-derived neurons. Three of these effects, one exerted at least in large part by adenosine and the others by agents still under study (one appears to be 5-HT), are described by Furshpan et al. (1986), Matsumoto et al. (in press), and D. Sah and S. G. Matsumoto (unpublished observations). This paper is concerned with evidence for secretion by these neurons of norepinephrine (NE) and acetylcholine (ACh). The physiological effects of the secretion of these two substances onto the myocytes (excitation and inhibition, respectively) were generally similar to those reported in vivo. The minimal latencies of the responses were short, probably due to the high density of innervation. ACh secreted by a neuron onto itself, at autapses, evoked fast nicotinic EPSPs. We have not detected autaptic effects attributable to the secretion of NE. A minority of the neurons were detectably only adrenergic or only cholinergic. The incidence of these transmitter states was strongly dependent on culture age and culture conditions; in a heterogeneous group of about 300 reasonably well-characterized neurons about 17% (12% of neonate-derived) were apparently purely adrenergic and about 10% (13% of neonate-derived) were apparently purely cholinergic.(ABSTRACT TRUNCATED AT 400 WORDS)