This study evaluated Streptococcus mutans biofilm adhesion on the surface of three composite resins (nanofilled, Filtek Z350, 3M ESPE, Salt Lake City, UT, USA; nanohybrid, Vit-1-escence, Ultradent Products, South Jordan, UT, USA; and microhybrid, Esthet X, Dentsply, Milford, DE, USA) following different finishing and polishing techniques. Sixty standardized samples (6 × 3 mm) of each composite were produced and randomly divided into three finishing and polishing treatments (n=20): 1) control group: composite resin surface in contact with Mylar matrix strips with no finishing or polishing performed, 2) Sof-Lex aluminum oxide disc technique (3M ESPE, and 3) carbide bur finishing and Astrobrush polishing technique (Ultradent). Half the samples of each group were incubated in human saliva for 1 hour, and all the samples were subjected to S mutans (ATCC 35688) biofilm development. The mean log of CFU/mL present in the S mutans biofilm was calculated, and data were statistically analyzed by three-way analysis of variance and the Tukey test (p<0.05). Human saliva incubation promoted a significant increase of bacterial adherence on all three of the composites' surfaces, regardless of the polishing treatment performed (p<0.05). Of the three, the nanofilled composite (Filtek Z350) had the lowest bacterial adherence with each of the finishing and polishing techniques despite the presence or absence of human saliva (p<0.05). Mylar matrix strips (control group) promoted the lowest bacterial adhesion on the surface of the microhybrid and nanofilled composites in the absence of human saliva.