Purpose: To evaluate the influence of previous acid etching on the bond strength of a self-etch universal adhesive system on sound and eroded enamel and dentin. Materials and Methods: Flat enamel and dentin surfaces were obtained from bovine incisors (n = 44) and divided into two groups according to the exposure to erosive conditions: sound (without erosion) and eroded (after erosive challenge with 0.3% citric acid, pH 2.6, 5 min, 4x/day, 5 days). Each group was then divided into two subgroups (n = 11) according to the application or not of 37% phosphoric acid before applying a universal self-etch adhesive. Single Bond universal adhesive (3M ESPE) was applied and composite blocks were built up using Amelogen Plus (Ultradent). Sticks (1 × 1 mm) were obtained and the microtensile bond strength (µTBS) was assessed using a universal testing machine at a speed of 1 mm/min. Data were analyzed with two-way ANOVA and Tukey's tests (p 0.05). Results: For enamel groups, significant differences were observed for erosive challenge (p = 0.034) and acid etching (p = 0.047), but not for the interaction between them (p = 0.182). The means ± SD (MPa) for the erosive conditions were: sound, 25.02 ± 5.82a; eroded, 28.45 ± 5.92b; and for acid etching: without, 24.89 ± 5.39a; with, 28.58 ± 6.23b (different superscript letters indicate significant differences). For dentin, both the erosive condition (p = 0.936) and acid pretreatment (p = 0.084), as well as their interaction (p = 0.107), did not affect the bond strength. Conclusion: The erosive challenge and previous acid etching significantly increased the bond strength of the tested universal adhesive to enamel but not to dentin. Keywords: tooth erosion, enamel, dentin, microtensile bond strength, adhesive system