This study tested the hypothesis that the hydration condition of the tooth during bonding or storage may affect the bond strength of a single-step, self-etching adhesive to dentin. Twelve extracted human third molars were transversally sectioned to expose flat, mid-dentin surfaces. Six teeth were dehydrated in ascending ethanol concentrations for several days while 6 teeth were kept hydrated in distilled water. The teeth were bonded with the single-step, self-etching adhesive One-Up Bond F (Tokuyama, Japan) according to directions. Resin composite buildups were constructed incrementally with Z 250 (3M ESPE). The bonded hydrated and dehydrated teeth were then divided into 2 subgroups and stored either in distilled water or in a dry condition for 24 h before being prepared for the microtensile bond strength test. Bonded beams of approximately 0.8 mm2 were tested in tension at 0.5 mm/min in a testing machine. Separate dentin disks were bonded with the adhesive and prepared for transmission electron microscopy (TEM). Bond strengths of specimens stored in water were: 25.0 ± 10.6 and 18.9 ± 9.1 MPa for hydrated and dehydrated teeth, respectively. Bond strengths of specimens stored dry were significantly higher (p 0.001): 46.1 ± 21.4 and 40.9 ± 19.1, for hydrated and dehydrated teeth, respectively. Two-way ANOVA showed that previous hydration condition of the teeth had no effect on the bond strength, regardless of the storage condition (p > 0.05). However, storage of the bonded teeth in water caused significant reductions in the bond strength, regardless of the previous hydration condition of the teeth. TEM showed massive silver nitrate impregnation in the adhesive layer of water-stored specimens. Postbonding water exposure had a highly significant effect on bond strength of the single-step, self-etching adhesive. Palabras clave: dentin bonding, water trees, permeability