Sevoflurane is a widely used obstetric general anesthetic, but the neurotoxic effects of late-pregnancy exposure to one minimum alveolar concentration ([MAC], 2.5%) of sevoflurane on offspring remain unclear. We investigated whether exposure to 2.5% sevoflurane during late pregnancy would affect offspring hippocampal neuronal development and neurocognitive function. On gestational day 18 (G18), rats were randomly treated with 2.5% sevoflurane in 50% oxygen for 1 (Sev × 1), 3 (Sev × 3), or 6 h (Sev × 6). The neuronal apoptosis rate and mature brain-derived neurotrophic factor (mBDNF) and postsynaptic density protein 95 (PSD-95) expression levels were measured in offspring hippocampi on postnatal day 1 (P1) and P35. Dendritic spine formation and cognitive function were examined on P35. The neuronal apoptosis rate was enhanced, and mBDNF and PSD-95 levels were reduced in the Sev × 3 and Sev × 6 groups on P1. mBDNF and PSD-95 levels were also decreased in the Sev × 6 group on P35. The error rate was elevated in the maze test, whereas dendritic spine density and long-term potentiation (LTP) were reduced in the Sev × 6 group on P35. To determine whether exposure to an enriched environment (EE) would ameliorate sevoflurane’s neurotoxic effects, offspring from another Sev × 6 group were exposed to either a standard environment (SE) or an EE. Lower error rates and greater dendritic spine densities and LTP were found in the Sev × 6 + EE vs. Sev × 6 + SE group. Collectively, we showed that exposing rats to 1 MAC sevoflurane for 3 h during late pregnancy increased neuronal apoptosis in neonates but did not impair neuronal development or cognitive function in juvenile rats, whereas a 6-h exposure impaired neuronal development and cognitive function in juvenile rats, effects that were attenuated by an EE.