The aim of this in vitro study was to optimise clinical parameters and the energy density of Er:YAG laser-conditioned dentin for class V fillings. Shear tests in three test series were conducted with 24 freshly extracted human third molars as samples for each series. For every sample, two orofacial and two approximal dentin surfaces were prepared. The study design included different laser energies, a thin vs a thick bond layer, the influence of adhesives as well as one-time- vs two-time treatment. The best results with Er:YAG-conditioned dentin were obtained with fluences just above the ablation threshold (5.3 J/cm(2)) in combination with a self-etch adhesive, a thin bond layer and when bond and composite were two-time cured. Dentin conditioned this way reached an averaged bond strength of 23.32 MPa (SD 5.3) and 24.37 MPa (SD 6.06) for two independent test surfaces while showing no statistical significance to conventional dentin adhesion and two-time treatment with averaged bond strength of 24.93 MPa (SD 11.51). Significant reduction of bond strength with Er:YAG-conditioned dentin was obtained when using either a thick bond layer, twice the laser energy (fluence 10.6 J/cm(2)) or with no dentin adhesive. The discussion showed clearly that in altered (sclerotic) dentin, e.g. for class V fillings of elderly patients, bond strengths in conventional dentin adhesion are constantly reduced due to the change of the responsibles, bond giving dentin structures, whereas for Er:YAG-conditioned dentin, the only way to get an optimal microretentive bond pattern is a laser fluence just above the ablation threshold of sclerotic dentin.