Recent anatomical evidence has cast doubt on the interpretation of the neural elements involved in past experiments in which efferents were electrically stimulated. To separate effects produced by medial olivocochlear (MOC) efferents from effects produced by lateral olivocochlear (LOC) efferents, MOC efferents were selectively stimulated by an electrode in the region of the MOC cell bodies in cats. For comparison, efferents were also stimulated with an electrode in the fourth ventricle (OCB stimulation, previously called COCB stimulation). MOC stimulation and fourth-ventricle OCB stimulation both produced qualitatively similar results bilaterally in that auditory-nerve compound action potential (N1) and endocochlear potential were reduced, and cochlear microphonic (CM) was increased. Both efferent-induced changes were affected in similar ways by changes in shock parameters, and were blocked by strychnine. At low sound levels, the decrease in N1 amplitude was approximately equivalent to a shift (decrease) in sound level but the change in N1 latency was not. The ratio of the CM increase to the N1 sound-level shift was independent of shock level or location. MOC stimulation typically produced an N1 sound-level shift of 11-16 dB in the contralateral ear and 4-7 dB in the ipsilateral ear. The ratio of these shifts almost equals the ratio of MOC neurons which had cell bodies on the stimulating-electrode side. Previous results reported by others with 'UOCB stimulation' now seem attributable to excitation of uncrossed MOC efferents rather than to excitation of uncrossed LOC efferents as previously thought. There is no effect reported in the literature or seen by us which can definitely be attributed to LOC neurons. Fourth-ventricle OCB stimulation typically produced an N1 sound-level shift in both ears of 19-22 dB which is approximately the sum of the crossed and uncrossed MOC shifts. Considering also that many uncrossed-MOC fibers course close to the midline (i.e. near the stimulating electrode), it seems likely that fourth-ventricle OCB stimulation excites both crossed and uncrossed MOC efferents. Referring to such stimulation in the cat as 'COCB stimulation' is therefore inaccurate and may lead to wrong conclusions about the functional role of various components of the olivocochlear fibers.