Recently, neurodegenerative diseases are more often linked with oral bacteria-driven infections. However, it is not clear how central nervous system pathologies root from infections of the oral cavity. Porphyromonas gingivalis (Pg) is a key-stone pathogen in periodontitis, a chronic inflammatory disease of the gingival and teeth-supporting structures. Many studies suggest that differences in infection severity and systemic pathologies are linked with different Pg strains and produced virulence factors. In this study, we used a zebrafish (Danio rerio) infection model, widely used in biomedical studies, including those of immune response to human pathogens. Importantly, this organism is genetically and physiologically similar to humans, and a plethora of zebrafish transgenic lines are available, allowing for in vivo real-time imaging of infection outcomes. We compared the ability of the two strains of Pg: fimbriated, non-encapsulated ATCC 33277 and non-fimbriated, encapsulated W83 to induce neuroinflammation using a zebrafish bloodstream infection model. Neuroinflammation was examined by gene expression of pro-inflammatory markers in the heads of infected larvae. We also compared the ability of these Pg strains to affect brain vascular and activate microglia. We found that, compared to the W83 strain, ATCC 33277 was more adherent to the endothelium, although it did not cause any vascular abnormalities. In contrast, Pg W83 was able to cross the brain vascular barrier and degrade brain vessels. Pg W83 also activated microglia, changing its morphology from highly ramified into ameboid. Zebrafish larvae provide invaluable insight into the complex mechanisms of infection-driven neuropathologies.