By use of the indirect immunofluorescence (IF) technique, radioimmunoassay (RIA) and in situ hybridization (ISH) histochemistry, the staining pattern, content and expression of calcitonin gene-related peptide (CGRP) in lumbar motoneurons of normal rats and rats subjected to sciatic nerve transection (SNT), ventral root transection (VRT), low thoracic spinal cord transection (SCT) alone or in combination with a subsequent SNT, as well as rats subjected to chemical lesioning of 5-hydroxytryptamine (5-HT) neurons by 5,7-dihydroxytryptamine (5,7-DHT), were studied. We here confirm that a large number of the lumbar motoneurons normally contain CGRP-like immunoreactivity (LI) and CGRP mRNA. SNT induced a transient increase in CGRP-LI, with a peak at days 2–5 after lesion, and normalized levels again after ∼2–3 weeks. Comparable results were obtained with IF and RIA. This increase is probably a consequence of increased CGRP synthesis, since a parallel up-regulation of CGRP mRNA levels was seen. A normalization of CGRP mRNA did not occur during the period studied, despite an apparent normalization of peptide levels after 2 weeks, and this may in turn be due to an increased turnover and/or release of CGRP. The up-regulation of CGRP is probably caused by the axon injury itself, since a similar cellular reaction with respect to CGRP was observed in motoneurons subjected to VRT. However, SNT, which also lesions dorsal root afferents and causes a decline in CGRP-LI in the dorsal horn, induced an increase in CGRP-LI in motoneurons on the contralateral side also. Thus, it may be that severance of dorsal root afferents and/or changes in reflex activity may also influence the production of CGRP in motoneurons. SCT, which severs all descending synaptic input to the motor nucleus and causes a paralysis of muscles innervated by motoneurons below the lesion, resulted in a marked decline in both content of CGRP-LI (IF and RIA) and expression of CGRP mRNA. However, treatment with 5,7-DHT, which lesions 5-HT neurons, including those giving rise to the bulbospinal serotoninergic pathway, did not cause any dramatic changes in motor behaviour but induced an increase in both motoneuron content of CGRP-LI and expression of CGRP mRNA. In rats first subjected to SCT, which depresses CGRP, followed 2 weeks later by SNT, we found a marked increase in both content of CGRP-LI (IF and RIA) and expression of mRNA coding for CGRP. In summary our results show that the cellular production of the CGRP peptide, normally expressed in motoneurons, is influenced in a complex way by motoneuron injury as well as changes in the afferent input. There also appear to be important differences in the expression of CGRP in small (gamma) and large (alpha) motoneurons as well as between motoneurons of different nuclei, in normal as well as axotomized rats.