The secretory pattern of GH is markedly sexually dimorphic in the adult rat, a phenomenon that becomes manifest around the time of pubertal development. This event is due partially to the pubertal rise in gonadal steroids. However, the fetal and neonatal sex steroid environments also play an important role in generating this sexual dimorphism. Hypothalamic mRNA levels of GH-releasing hormone (GHRH) and somatostatin (SS), two neuropeptides implicated in the control of GH release, are sexually dimorphic in both neonatal and adult animals and, at least in the adult animal, are responsive to modulation by sex steroids. In this study, we examined the effects of neonatal testosterone on the number of GHRH and SS neurons in the adult hypothalamus as well as its effects on the responsivity of these neurons to later increases in sex steroids. To address these questions, male rats were either castrated or sham castrated on the day of birth (P0); these animals, along with intact females, received an injection of either testosterone or vehicle. At 60 days of age, half of each group received a Silastic capsule containing testosterone, and half received a sham implant. Growth rates were monitored throughout the study. At 75 days of age, animals were killed, and in situ hybridization to detect GHRH and SS mRNA containing neurons was performed. The number of GHRH neurons in the arcuate nucleus and ventromedial hypothalamus and the number of SS neurons in the periventricular nucleus and paraventricular nucleus were counted. Using a computerized image analysis system, GHRH and SS mRNA signal levels in individual neurons were also measured. Both neonatal and adult steroid treatments significantly increased growth rates. Those animals exposed to neonatal testosterone had significantly more detectable GHRH neurons than those that received only vehicle [P < 0.0001, by analysis of variance (ANOVA)]. Neonatal testosterone treatment had no effect on GHRH mRNA levels. Adult testosterone treatment, while having no effect on GHRH neuron numbers, stimulated GHRH mRNA levels in both males and females (P < 0.0001, ANOVA), but the magnitude of the increase depended upon whether the animal had been exposed to testosterone during the neonatal period. In contrast, the number of SS neurons was not affected by either steroid treatment. However, both treatments modulated SS mRNA levels (P < 0.0001, by ANOVA), with neonatal testosterone treatment alone resulting in significantly higher levels of SS mRNA in the adult animal. Adult testosterone treatment also significantly increased SS mRNA levels, and this was independent of previous exposure to sex steroids.(ABSTRACT TRUNCATED AT 400 WORDS)