Background/Objectives: Maternal obesogenic diets may produce persistent developmental alterations in offspring, with response patterns that may differ between males and females. We investigated whether propolis supplementation from preconception through lactation modified metabolic and hepatic outcomes in offspring exposed to a maternal cafeteria diet. Methods: Thirty-two Wistar rat dams received a control diet (CON), cafeteria diet (CAF), or either diet with propolis in drinking water (CONP and CAFP; target exposure, 200 mg/kg/day). One male and one female per litter (n = 8/sex/group) received standard chow after weaning. This design isolated maternal exposure because offspring received no propolis after weaning. At 18 weeks, metabolic and inflammatory biomarkers, liver histopathology, and hepatic lipid-metabolism gene expression were evaluated using groupwise tests and two-way ANOVA. Results: Final body weights were comparable across groups. In males, maternal diet affected glucose (p = 0.034) and HOMA-IR (p = 0.023), with a diet × propolis interaction for glucose (p = 0.042). CAF males had higher TNF-α than CAFP and CONP males. In females, diet × propolis interactions were detected for insulin, total cholesterol, triglycerides, and leptin (all p ≤ 0.049); CAF females had higher triglycerides than CON and CAFP females. Hydropic degeneration was greater in CAF dams than all other groups and in CAF than CONP male offspring; female histopathology did not differ. Male hepatic gene expression did not differ among groups, whereas female PPAR-β and FADS2 showed propolis main effects (p = 0.046 and p = 0.009, respectively). Conclusions: Maternal cafeteria-diet exposure was associated with persistent metabolic and hepatic alterations despite comparable offspring body weights, with different response patterns observed in the separate male and female analyses. Maternal propolis supplementation selectively modified some inflammatory, lipid-related, and histopathological outcomes; however, these effects were heterogeneous and did not consistently normalize the cafeteria-associated phenotype. Overall, the findings support partial, endpoint-specific, and context-dependent modulation rather than uniform metabolic or hepatic protection.