The aim of this study was to evaluate the influence of the LED curing unit and selective enamel etching on dentin microtensile bond strength (mu TBS) for self-etch adhesives in class I composite restorations. On 96 human molars, box-shaped class I cavities were made maintaining enamel margins. Self-etch adhesives (Clearfil SE - CSE and Clearfil S-3 - S3) were used to bond a microhybrid composite. Before adhesive application, half of the teeth were enamel acid-etched and the other half was not. Adhesives and composites were cured with the following light curing units (LCUs): one polywave (UltraLume 5 - UL) and two single-peak (FlashLite 1401 - FL and Radii Cal - RD) LEDs. The specimens were then submitted to thermomechanical aging and longitudinally sectioned to obtain bonded sticks (0.9 mm(2)) to be tested in tension at 0.5 mm/min. The failure mode was then recorded. The mu TBS data were submitted to a three-way ANOVA and Tukey's (alpha = 0.05). For S3, the selective enamel-etching provided lower mu TBS values (20.7 +/- 2.7) compared to the non-etched specimens (26.7 +/- 2.2). UL yielded higher mu TBS values (24.1 +/- 3.2) in comparison to the photoactivation approach with FL (18.8 +/- 3.9) and RD (19.9 +/- 1.8) for CSE. The two-step CSE was not influenced by the enamel etching (p a parts per thousand yenaEuro parts per thousand 0.05). Enamel acid etching in class I composite restorations affects the dentin mu TBS of the one-step self-etch adhesive Clearfil S-3, with no alterations for Clearfil SE bond strength. The polywave LED promoted better bond strength for the two-step adhesive compared to the single-peak ones.27611751182