The social reproductive inhibition evident in female common (Cryptomys hottentotus hottentotus) and highveld (Cryptomys hottentotus pretoriae) mole-rats was investigated, focussing on intra- and inter-subspecies differences in gonadotropin¬releasing hormone (GnRH) neuroanatomy and neuroendocrinology and potential mechanisms mediating social infertility through the suppression of luteinizing hormone (LH) and/or GnRH synthesis or release. The molecular structure of the GnRH peptide present in highveld mole-rat tissue was determined. Messenger ribonucleic acid (mRNA) coding for GnRH was isolated from hypothalamic tissue of a female highveld mole-rat. The nucleotide sequence of the transcribed complimentary deoxyribonucleic acid (cDNA) predicts amino acid sequences that indicate a potential polymorphism in the mature peptide. These sequences differ from classic 'mammalian' GnRH by either one or two amino acid positions. One form of mole-rat GnRH is identical to the unique GnRH expressed by guinea pigs, while the second form differs from guinea pig GnRH in one amino acid position. These results indicate that classic 'mammalian' GnRH is not the chief structural form of GnRH in the mole-rat. In both the common and highveld mole-rat, GnRH structures were loosely distributed along the septo-preoptico-infundibular pathway. Highveld mole-rats consistently had more GnRH cell bodies and higher levels of the peptide than common mole-rats. The subspecies also differed with regards to the relative distribution of GnRH cell bodies in the preoptic area and the medio-basal hypothalamus. Reproductive status and season did not influence GnRH neuroanatomy (numbers and morphology of cell bodies, distribution of GnRH structures and density of GnRH staining in the median eminence) or endogenous brain GnRH content in male or female common mole-rats. Likewise reproductive and non-reproductive male highveld mole-rats did not differ with regards to the GnRH neuroanatomical and neuroendocrinological parameters investigated. Dominant and subordinate female highveld mole-rats had similar numbers of GnRH cell bodies, and did not differ with regards to the distribution of GnRH structures. Brain GnRH levels was however significantly lower in reproductive females compared to non-reproductive females. Similarly, GnRH staining in the median eminence was significantly less intense in reproductive compared to non-reproductive females, indicating reduced release in the latter. Excess peptide most likely accumulates in the nerve terminals and the cell bodies. Endogenous opioid peptides (EOPs) are tentatively considered to playa role in the socially-induced infertility in non-reproductive female highveld mole-rats. Intact reproductive females and intact and ovariectmized non-reproductive females were treated with the opioid antagonist naloxone, alone and in different combinations with exogenous GnRH, to establish the role of EOPs and gonadal hormones in the reduced GnRH release evident in subordinate female highveld mole-rats. Naloxone treatment failed to alter plasma LH levels in any of the intact groups, while ovariectomized non-reproductive females showed a significant response. The pituitary response to a GnRH challenge was not influenced by either a single naloxone administration or longer term naloxone-priming regimens. In the absence of the negative feedback effects of gonadal steroids following ovariectomy, non¬reproductive females showed GnRH-chalienged plasma LH levels similar to that seen in intact reproductive females, both being significantly higher compared to intact non-reproductive females. Steroid-dependent mechanisms altering GnRH secretion and/or pituitary sensitivity to the endogenous peptide may therefore playa role in the social reproductive regulation displayed by the highveld mole-rat. EOPs may be indirectly involved in steroid negative feedback suppression of GnRH. In conclusion, the GnRH neuroanatomy and neuroendocrinology of female common mole-rats provide no evidence for a physiological component to the behavioural reproductive restraint in subordinate females due to incest avoidance. Non¬reproductive female highveld mole-rats are, however, temporarily infertile due to pituitary insensitivity to GnRH coupled with suppressed hypothalamic GnRH release potentially mediated by steroid-dependent mechanisms.