The role of estrogen in lordosis-inhibiting systems in the lateral septum or the dorsal raphe nucleus was investigated in female and male rats. Ovariectomized rats received implantation of 22-gauge guide cannulae to the bilateral or right side of the lateral septum (LS and rLS, respectively), the dorsal raphe nucleus (DRN) or bilateral cortex (CX). In castrated male rats, bilateral implantations of the cannulae to the LS were carried out (mLS). Three behavioral tests in total were performed at 2-week intervals. In the first test, all animals were subcutaneously injected with 1.5 microg/kg estradiol benzoate (EB). Forty-four hours after EB, 0.5 mg progesterone (P) was injected and a behavioral test was started 4 h after P. These hormonal regimes were used in all tests. In the second test, 2 h before EB injection, 27-gauge cannulae filled with estradiol (E2) were inserted into the DRN, LS or CX through the guide cannulae and were kept there for 4 h. In the third test, cholesterol was implanted instead of E2 into these areas. In the first test, most females showed low levels of lordosis quotient (LQ) and most males showed no lordosis. In the second test, mean LQs in the LS or rLS groups of females increased but not in the DRN and CX groups. In the mLS group no increase of LQ was observed. When cholesterol was implanted in the third test, mean LQs in all groups were as low as in the first test. These results suggest the possibility that estrogen releases the inhibition when it acts on the LS, but not on the DRN female rats. On the other hand, inhibition in the male LS may not be released by the direct action of estrogen.