Carbon-based nanomaterials are being increasingly used, demanding strong information to support their safety in terms of human health. As ingestion is one of the most important exposure routes in humans, we have determined their potential risk by using an in vitro model simulating the human intestinal barrier and evaluated the effects of both graphene oxide (GO) and graphene nanoplatelets (GNPs). A coculture of differentiated Caco-2/HT29 cells presenting inherent intestinal epithelium characteristics (i.e. mucus secretion, brush border, tight junctions, etc.) were treated with GO or GNPs for 24 h. Different endpoints such as viability, membrane integrity, NPs localization, cytokines secretion, and genotoxic damage were evaluated to have a wide view of their potentially harmful effects. No cytotoxic effects were observed in the cells that constitute the barrier model. In the same way, no adverse effects were detected neither in the integrity of the barrier (TEER) nor in its permeability (LY). Nevertheless, a different bio-adhesion and biodistribution behavior was observed for GO and GNPs by confocal microscopy analysis, with a more relevant uptake of GNPs. No oxidative damage induction was detected, either by the DCFH-DA assay or the FPG enzyme in the comet assay. Conversely, both GO and GNPs were able to induce DNA breaks, as observed in the comet assay. Finally, low levels of anti-inflammatory cytokines were detected, suggesting a weak anti-inflammatory response. Our results show the moderate/severe risk posed by GO/GNPs exposures, given the observed genotoxic effects, suggesting that more extensive genotoxic evaluations must be done to properly assess the genotoxic hazard of these nanomaterials.