Abstract: This study compared the turnover of GABA neurons in different brain areas of the male rat and examined the effect of castration on GABA turnover in regions of the brain associated with the control of gonadotropin secretion. To estimate GABA turnover, GABA was quantified by HPLC in microdissected brain regions 0,30,60,90, and 120 min after inhibition of GABA degradation by aminooxyacetic acid (100 mg/kg, i.p.). GABA accumulation was linear in all areas for 90 min (p < 0.01), and GABA turnover was estimated as the slope of the line formed by increased GABA concentration versus time, determined by linear regression. There was considerable regional variation both in the initial steady-state concentrations of GABA and in the rates of GABA turnover. Of 10 discrete brain structures, GABA turnover was highest in the medial preoptic nucleus and lowest in the caudate nucleus. Turnover times in the terminal fields of known GABAergic projection neurons ranged sevenfold, from 2.6 h in the substantia nigra to 0.4 h in the lateral vestibular nucleus. The effect of castration on GABA turnover in 13 microdissected brain regions was investigated by measuring regional GABA concentrations before and 30 min after injection of aminooxyacetic acid in intact rats or 2 or 6 days postcastration. Following castration, steady-state GABA concentrations were increased, and GABA turnover decreased in the diagonal band of Broca, the medial preoptic area, and the median eminence. GABA turnover increased in the medial septal nucleus and was unaffected in the cortex, striatum, and hindbrain. These results are consistent with the hypothesis that testosterone negative-feedback control of luteinizing hormone-releasing hormone involves steroid-sensitive GABAergic neurons in the rostral and medial basal hypothalamus.