Objective To analyze the disparities in social, emotional, and cognitive behaviors between male and female individuals with autism spectrum disorder (ASD) resulting from various exposure methods of valproic acid (VPA), and to establish a foundation for selecting experimental animal gender and induction method in the VPA-induced ASD model. Methods At 12.5 days of gestation, the experimental mice received an intraperitoneal injection of 600 mg/kg VPA, while the control group received an equivalent volume of normal saline. At 14 days post-birth, the experimental mice were administered a subcutaneous injection of 400 mg/kg VPA, with the control group receiving the same volume of normal saline. Additionally, at 14 days post-birth, the experimental mice were given an intraperitoneal injection of 400 mg/kg VPA, while the control group again received an equivalent volume of normal saline. The mortality and abortion rates of the experimental mice post-administration were monitored; through a series of three-box social behavior experiments, self-grooming assessments, open field tests, Y-maze trials, and water maze evaluations, the impact of various VPA exposure methods on the social, emotional, and cognitive behaviors of male and female mice was examined. Results Following VPA injection at 12.5 days of gestation, the mortality rate among pregnant rats was 50%, and the abortion rate was 25%. No fatalities occurred following subcutaneous or intraperitoneal administration of VPA postnatally. The outcomes of the three-box social experiment, self-grooming experiment, and open field experiment indicated that, in comparison to the control group, both male and female VPA mice displayed social impairments, novelty-induced social deficits, repetitive stereotyped behaviors, and anxiety-like behaviors, irrespective of the injection method employed. The outcomes of the Y-maze spontaneous alternation experiment and the water maze experiment indicated that male mice administered VPA prenatally and those receiving subcutaneous VPA postnatally exhibited cognitive deficits, whereas male mice injected with VPA intraperitoneally postnatally demonstrated a propensity for cognitive impairment without statistical significance; additionally, female mice exposed to VPA via various methods did not exhibit cognitive impairments. Conclusions The mortality and abortion rates of pregnant mice exposed to VPA before birth were higher, while no deaths occurred after subcutaneous or intraperitoneal injection of VPA after birth. Male mice subjected to postnatal subcutaneous VPA and prenatal VPA administration had autism-like characteristics, including diminished social interaction, repeated stereotyped actions, heightened anxiety, and cognitive impairment. However, the female mice were unstable. Postnatal intraperitoneal administration of VPA in mice led to moderate social and social novelty deficits, slight anxiety-like behaviors, and no significant impairment in learning abilities.