Plasma testosterone (T) was measured in control male and female rats on gestational days 16, 17, 18, 19, and 20 and on days 17-20 in males from dams who were fed ethanol and/or were stressed during pregnancy. Circulating T in control males showed an earlier rise, yielding a longer period of prenatal T elevation, than was reported previously (Endocrinology 106 (1980)306). Compared to control males, exposure to alcohol-alone augmented T on days 18 and 19, stress-alone attenuated prenatal T, and the combination of stress and alcohol completely blocked the normal rise in T between days 17 and 18. When these prenatal alterations in T are viewed along with effects these same treatments have on the postparturient T surge (Horm. Behav. 41 (2002) 229), a possible explanatory mechanism emerges for the uniquely different behavioral patterns of sexual behavior differentiation induced in males by prenatal exposure to alcohol, stress, or both factors. Whereas the potential for feminine behavior is retained to the extent that either the prenatal or the neonatal T surge is attenuated, the male potential is more sensitive to reductions in the fetal surge and is maximally disrupted if both the prenatal and the postparturitional T surges are suppressed.