Maternal infection during pregnancy is a risk factor for some psychiatric illnesses of neurodevelopmental origin such as schizophrenia and autism. In experimental animals, behavioral and neuropathological outcomes relevant to schizophrenia have been observed in offspring of infected dams. However, the type of infectious agent used and gestational age at time of administration have varied. The objective of the present study was to compare the effects of prenatal challenge with different immune agents given at different time windows during gestation on behavioral outcomes in offspring. For this, pregnant rats were administered bacterial endotoxin (lipopolysaccharide, LPS), the viral mimic polyinosinic: polycytidylic acid (poly I:C), or turpentine, an inducer of local inflammation, at doses known to produce fever, at three different stages in pregnancy: embryonic day (E)10-11, E15-16 and E18-19. Prepulse inhibition of acoustic startle (PPI) was later measured in male adult offspring. PPI was significantly decreased in offspring after prenatal LPS treatment at E15-16 and E18-19. Intramuscular injection of pregnant dams with turpentine at E15-16 also decreased PPI in adult offspring. Maternal poly I:C administration had no significant effect on PPI in offspring. In contrast to prenatal LPS exposure, acute LPS administration to naive adult males had no effect on PPI. Thus, prenatal exposure both to a systemic immunogen and to local inflammation at brief periods during later pregnancy produced lasting deficits in PPI in rat offspring. These findings support the idea that maternal infection during critical windows of pregnancy could contribute to sensorimotor gating deficits in schizophrenia.