To examine the role of progesterone (P) and its 5α-reduced metabolite, the neurosteroid 5α-pregnan-3α-ol-20-one (3α,5α-THP), in endogenous variations in ictal activity rats were tested for kainic acid-induced seizures in different hormonal milieu. Corresponding plasma and central P and 3α,5α-THP levels were measured. Cycling Long-Evans rats in estrus and proestrus had seizures of significantly shorter duration and more central and plasma 3α,5α-THP and P than animals in metestrus or diestrus. Females with luteal functioning had seizures of significantly shorter duration and increased central and plasma 3α,5α-THP and P compared to animals that recently had luteal functioning discontinued. Pregnant rats had significantly shorter seizures and greater central and plasma 3α,5α-THP and central P than animals tested 1–2 days postparturition. In all test paradigms, seizure activity was increased in animals that had decreased 3α,5α-THP or P; overall, central 3α,5α-THP was more inversely related to ictal activity than central P or plasma P and 3α,5α-THP. To investigate a causal relationship between 3α,5α-THP and seizures, a 5α-reductase inhibitor, finasteride, or vehicle was administered to pregnant rats. Finasteride administration significantly decreased central and plasma 3α,5α-THP, but had no significant effect on plasma or central P of pregnant rats. Finasteride, but not vehicle administration, to pregnant rats significantly increased seizure duration. These findings support the hypothesis that variations in seizure threshold over endogenous hormonal milieu may be related to endogenous 3α,5α-THP. Of all of the endocrine conditions, seizure durations were greatest in diestrus animals; this group did not experience the lowest or the greatest decrease in 3α,5α-THP concentrations; however, of all of the endocrine conditions, cycling rats experienced the most rapid cycles of 3α,5α-THP variation. This suggests that cycles of endogenous variations in 3α,5α-THP may influence seizure threshold.