This study evaluated the cariostatic effect of antibacterial self-etching adhesive systems, by means of an in vitro bacterial caries model. Seventy-five prepared bovine slabs were randomly divided into groups (n=15): (1) unbonded composite, no carious challenge (UNB-NC); (2) unbonded composite, carious challenge (UNB-C); (3) Clearfil SE Bond, no antibacterial agent (CSE); (4) Protect Bond, containing MDPB and fluoride (PB); and (5) Reactmer Bond, fluoride-releasing (RB). All preparations were restored with Filtek Z-250. Groups (2)-(5) were submitted to a medium containing Streptococcus mutans (ATCC-- 25175) for 5 days, and Group (1) was kept in a noninoculated medium. Insoluble polysaccharides present in tooth biofilms were quantified, Knoop hardness (KHN) was measured on the enamel adjacent to restorations, and standard 35-mm polarized light photomicrographs were taken as illustrations. Polysaccharide and Knoop hardness results were analyzed with the use of ANOVA, with a split-split-plot statistical design for KHN. Except for Group (1), all groups showed similar caries formation. Biofilm over PB restorations showed the smallest amounts of polysaccharides (14.37 microg/mg), and CSE showed the highest amounts (20.87 microg/mg). All self-etching systems tested were unable to inhibit secondary caries in a bacterial model simulating a high caries challenge, even though there was reduced glucan synthesis provided by the adhesive system containing MDPB and fluoride.