In this study, the cytotoxicity of different combinations of contemporary resin‐based restoratives (adhesives, composites, luting agents) against human keratinocytes (HaCaT) was evaluated under two conditions, whether materials were applied to dentin or not. Adhesives (3‐step etch‐and‐rinse/3ER: OptiBond FL; 2‐step self‐etch/2SE Clearfil SE Bond; Single Bond Universal/UNI), composites (conventional composite resin/CCR: Filtek Z350XT; flowable/FCR: Filtek Z350XT Flow; self‐adhesive composite resin/SACR: Dyad Flow), and luting agents (conventional luting agent/CLA: Variolink‐II; self‐adhesive luting agent/SLA: RelyXU200) were combined according to their clinical use. Eluates from polymerized specimens applied to dentin were placed in contact with cells grown for 1 and 7 d. The controls were defined by cells without material contact. Cell viability was determined using MTT [3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide)] assay. C=C conversion was investigated using Fourier‐transform infrared spectroscopy. After 1 d of incubation, when dentin was not present, 2SE yielded the highest cell viability, whereas 3ER, UNI, and SACR showed higher cell viability in the presence of dentin. After 7 d, when dentin was absent, 2SE and CLA achieved significantly higher cell viability. The presence of dentin resulted in a drastically higher cell viability for all materials, except 2SE and CLA. UNI had the lowest C=C conversion. The presence of dentin was a significant factor, which resulted in higher cell viability than what was seen for the material specimens per se. All materials resulted in a lower viability of HaCaT than what was seen under the no‐material control conditions, with effects mainly limited to the first 24 h.