AIM: Eugenol is claimed to interfere with the polymerization of composite resins and to affect shear bond strengths of dentine-bonded composite restorations. Eugenol-based sealers are used during root canal treatment in teeth that may require build-ups or extensive restorations. Unfortunately, the adverse effect of eugenol has mostly been assessed in shear bond tests and this variable may be clinically inappropriate. The current study evaluate the effect of eugenol in a eugenol-based endodontic sealer on marginal adaptation of composite resin restorations with and without thermo-mechanical stress. METHODOLOGY: Thirty Class V cavities with half of the cavity margins in dentine and half of the cavity margins in enamel were prepared in extracted human premolars and then divided into 5 groups. Teeth in group 1 were not contaminated with eugenol. Cavities in groups 2-4 were contaminated with a eugenol-based sealer (Tubli-Seal, Kerr) and cleaned with: (i) sandblasting, (ii) bur finishing alone or (iii) bur finishing combined with swabbing with alcohol. These 4 groups were temporized for 6 weeks (Ketac, ESPE), whilst group 5 received no eugenol, was not temporized and acted as the control. Cavities were restored with Tetric Ceram using Syntac Classic (Vivadent) as a dentine-bonding agent. The specimens were then thermo-mechanically stressed. The percentages of marginal adaptation in dentinal and enamel margins were assessed in a SEM at x200 magnification before and after stress using a replica method. Mean percentages of marginal adaptation were calculated and compared using a two way-ANOVA (influence of eugenol and cleaning procedure) and non-parametric tests. RESULTS: Marginal adaptation in enamel exceeded 95% and 92% in all specimens before and after thermo-mechanical stress, respectively. Before stress, marginal adaptation in dentine ranged from 92.3 +/- 7.9% to 95.7 +/- 6.2% in groups 1-5. After stress, the percentage of marginal adaptation in dentine decreased significantly overall (P < 0.05, range 39.8 +/- 21.1% to 82.9 +/- 13.7%). The effect of contamination with eugenol was not significant (P > 0.05): in contrast, there was a significantly beneficial effect when the entire cavity was finished with burs (P < 0.01). CONCLUSIONS: Under the conditions of this study, the eugenol containing sealer did not significantly impair marginal adaptation in dentinal margins of mixed Class V restorations when bur finishing was employed before placement. However, no other cleaning method was sufficient to produce acceptable figures of marginal adaptation.