Specific effects of stimulating different parts of the reticulospinal (RS) system on the spinal locomotor pattern are described in lamprey. In the in vitro brain stem and spinal cord preparation, microstimulation in different areas of the reticular formation was performed by ejecting a small amount of -glutamate from a micropipette. These areas were distributed over the four reticular nuclei of the brain stem: the mesencephalic reticular nucleus (MRN) and the anterior, middle and posterior rhombencephalic reticular nuclei (ARRN, MRRN, and PRRN, respectively). To prevent synaptic spread of excitation within the brain stem, the synaptic transmission was blocked by using a low Ca2+, high Mn2+ physiological saline in the brain stem pool. "Fictive" locomotion was evoked by applying N-methyl--aspartate (NMDA) to the spinal cord. Rhythmical discharges of motoneurons were recorded bilaterally in the midbody area, from the ventral roots that had been subdivided in dorsal and ventral branches, supplying the dorsal and ventral part of the myotome, respectively. Two major effects of brain stem stimulation were elicited: a change in the frequency of the locomotory rhythm and an induction of asymmetry (left/right, dorsal/ventral) in the segmental motor output. Approximately 50% of the stimulated sites evoked a change in locomotor frequency. In the PRRN almost all effective sites evoked an increase in frequency (10-50%). In the other nuclei, increase and decrease (10-30%) were observed equally frequently. Most of the stimulated sites (50-80%) in any reticular nucleus evoked asymmetry in the segmental motor output. Distortion of the segmental output symmetry was classified into eight categories by comparing the intensity of locomotor bursts in the dorsal and ventral branches of the two ventral roots, ipsilateral and contralateral to the stimulated side. These categories differed in the direction of the body flexion, which would be evoked during normal swimming: ipsilateral (I), contralateral (C), dorsal (D), ventral (V), ipsilateral and dorsal (ID), ipsilateral and ventral (IV), contralateral and dorsal (CD), and contralateral and ventral (CV). The different categories were not equally represented in each nucleus and across the nuclei. The most pronounced categories for each nucleus were as follow. In MRN: I (33%); ARRN: C (44%); MRRN: rostral part, I (36%) and caudal part, CV (42%); and PRRN: rostral part, I (40%) and caudal part, IV (35%). Other categories were also present but less common in each nucleus. To examine if the effects of brain stem stimulation were uniform along the spinal cord, recordings were performed from distal parts of the cord. Stimulation of a given point in the brain stem produced similar pattern of effects in 59% of cases and different patterns in 41% of cases. The main conclusion of the present study is that the proportion of RS neurons with different influences on the spinal locomotor network differs significantly among different parts of the reticular formation of the lamprey. The specificity of RS influences may represent a basis for modifications of the segmental locomotor output necessary for the control of equilibrium and steering during locomotion.