The rate and direction of neurite growth have been shown in a number of studies to be determined by the distribution of adhesive sites on the growth cone. Recent evidence showing that the application of extrinsic electric fields can redistribute membrane molecules and alter both the rate and direction of neurite growth have raised the question whether endogenous electric fields might be produced by steady currents in growth cones. To investigate this question, we have devised a novel circularly vibrating microprobe capable of measuring current densities in the range of 5 nA/cm2 (near the theorectical limit of sensitivity), with a spatial resolution of 2 micron. The design of this device and the development of a novel algorithm for computing current vectors on-line is described. Using this probe we have found that cultured goldfish retinal ganglion cell growth cones generate steady inward currents at their tips. The measured currents, in the range of 10-100 nA/cm2, appear to flow into the filopodia at their tips and back outward near the junctures of the filopodia and the growth cone. The currents appear to be produced only during active growth. Ion substitution experiments support the conclusion that the majority of this current is carried by Ca2+ ions, which we postulate flow through a population of activated voltage-sensitive Ca2+ channels located on the filopodial tips. Calculation of the transmembrane current density (4 X 10(-6) nA/cm2) leads to an estimate of channel density (10 channels/micron2) in close agreement with the measured density of Ca2+ channels in other systems. The assumption that calcium channel proteins are conveyed to nerve terminals by active transport, whereas sodium channel proteins are conveyed passively by a slower somatofugal diffusion process [Strichartz et al, 1984], would explain why developing neurons tend to display Ca2+-sensitive electrogenesis at their growing tips, and Na+-sensitive action potentials later in development. In order to gain some insight into the possible role of these steady growth currents, we estimated the membrane depolarization and axial voltage gradient they produce. It is likely that the currents produce sufficient membrane depolarization (approximately equal to 4 mV) to cause autogenous activation of ion channel permeabilities. Similarly, the axial voltage gradient (approximately equal to 4 mV/cm) would be expected to move intracytoplasmic vesicles by electrophoresis at a rate (20-40 microns/hr) very close to that at which the filopodia are observed to grow.(ABSTRACT TRUNCATED AT 400 WORDS)