PURPOSE To evaluate the influence of different pretreatments on the microtensile bond strength (µTBS) of an etch-and-rinse adhesive to eroded dentin. MATERIALS AND METHODS Thirty-six human teeth were ground down to their dentin layer and randomly divided into six groups (G1-G6; n = 6), G1 being the control group. Only in the test groups (G2-G6) were samples subjected to erosion using citric acid (pH 2.6) 10 x 2 min per day for five days. Between the erosive attacks, samples were stored in artificial saliva. After pretreatment - none (G1); none (G2); 2% chlorhexidine (30 s) (G3); prolonged primer application (1 min) (G4); roughening with a diamond bur (G5) and 10.5% NaOCl (1 min) (G6) - the adhesive OptiBond FL was applied. After the application of composite, samples were stored in water (7 d) and µTBS was determined. Data were evaluated using one-way ANOVA and Dunnett-T post-hoc test (p < 0.05). RESULTS Eroded dentin without pretreatment (G2) resulted in significant reduction of µTBS compared with uneroded dentin (G1). µTBS after pretreatment with a diamond bur (G5) or NaOCl (G6) was not significantly different from that of the uneroded control group (G1). µTBS after pretreatment with chlorhexidine (G3) or with prolonged primer application (G4) was significantly lower than in the uneroded control group (G1), and not significantly different from the eroded control group (G2). CONCLUSION The present data suggests that µTBS to eroded dentin pretreated with bur abrasion or NaOCl is similar to the µTBS to sound, uneroded dentin.