Chronic stimulating microelectrodes fabricated from platinum-30% iridium (Pt-30%Ir) or activated iridium were implanted in assemblies of three in the left sensorimotor cortex of the cat and pulsed continuously at currents of 10 to 320 microA (100 to 3200 microC/cm2 X ph, 2 to 64 nC/ph) for periods of 24 h or for 23 h/day for 7 days. The microelectrodes had beveled tips with uninsulated geometric surface areas of 20 X 10(-6) cm2. Neuronal activity evoked by the focal stimulation was monitored by recording compound action potentials from the ipsilateral pyramidal tract. By this criterion neuronal activation thresholds were 5 to 15 microA (50 to 150 microC/cm2 X ph, 1 to 3 nC/ph) for both types of electrodes. Histologic evaluations of tissue surrounding the electrode tips were carried out by either light or electron microscopy. No neural damage was induced by 24 or 161 h of pulsing using either type of electrode at currents of 10 to 80 microA. Neural damage attributable to electrical stimulation per se was observed in a few sites pulsed with 320 microA (3200 microC/cm2 X ph, 64 nC/ph, 16 A/cm2) with Pt-30%Ir but not activated iridium electrodes of the same size. Electrode dissolution appears to be best correlated with charge density and current density. Dissolution of the Pt-30%Ir microelectrode tip was observed by scanning electron microscopy at charge densities as low as 200 microC/cm2 X ph (1 A/cm2), whereas erosion of activated iridium microelectrodes occurred only at the highest charge and current densities (3200 microC/cm2 X ph, 16 A/cm2). Thus, the activated iridium electrode is superior to Pt-30%Ir for chronic stimulations, from the standpoint of electrode tip stability, because with the former, in contrast to the alloy, detectable erosion occurred only at an intensity well above that required for activation of nearby neurons.