Current data on the neurotrophic effects of steroid hormones suggest that, in brain and spinal cord regions containing receptor systems, steroids act at the level of RNA and protein synthesis to effect metabolic changes associated with nerve-cell survival, elaboration/maintenance of dendritic and axonal processes, synaptogenesis, and neurotransmission. While many of these effects appear to be associated with the neuroanatomical systems involved in the endocrine and behavioral aspects of reproduction, evidence does exist for similar neurotrophic effects outside the reproductive sphere. Both estrogens and androgens appear to exert this stimulatory, growthlike effect on target neurons. The effects of progesterone are not discussed in this review because relatively little information is available regarding the independent effects of progesterone on the brain. We have just completed a study (Jones et al., 1987b) which suggests that progesterone may act independently in the brain to affect protein synthesis. A number of conclusions concerning the mechanism of steroid action in producing trophic effects on neurons can be drawn. First, the time course of hormone action is similar to that found in nonneural target tissue, such as the uterus. Second, steroid hormones act on neurons through receptor-mediated genomic activation. Third, this effect on the genome appears to be at the level of both transcription and translation. Fourth, there is brain-region specificity in the gene products resulting from steroid hormone administration. Finally, short-term exposure to estrogens or androgens generally results in an anabolic response within target neurons. The brain and spinal cord, injured either by disease or by experimentally induced trauma, is responsive in a reparative manner to exogenous and/or endogenous gonadal steroid hormones. The mechanism underlying this therapeutic role of steroids on damaged neurons is not known but has been postulated to involve direct action of steroid hormones or target neurons. It has been hypothesized that two diseases, Alzheimer's and ALS, may be related to steroid hormone/receptor deficiencies. In this regard, Appel (1981) has suggested that putative "neurotrophic hormones" acting at the synapse may be critical in maintaining the neural networks affected in ALS, Alzheimer's disease, and parkinsonism. Extending that hypothesis to include direct action of such putative hormones within the cell body and at the level of the genome, the evidence presented in this discussion would argue that possible candidates could be gonadal steroids.(ABSTRACT TRUNCATED AT 400 WORDS)