Purpose: To evaluate in vitro the effect of clinical primer application errors on human dentin permeability. The null hypothesis is that there is no difference in dentin permeability reduction with the use of a total-etch three-step (TE3) dentin adhesive system among the treatment groups. Materials and Methods: Seventy extracted noncarious human third molars were sectioned to obtain 0.7-mm-thick midcoronal dentin disks. The specimens were randomly assigned to 5 groups and treated with a total-etch three-step (TE3) ethanol/water-based dentin adhesive system, Optibond FL (Kerr; Orange, CA, USA), according to manufacturer's instructions using 4 simulated application errors: short application of primer by immediate drying; no primer application; no drying of primer; aggressive drying of primer. Permeability of dentin was measured as fluid filtration at baseline and after adhesive polymerization. Kruskal-Wallis and Tukey's post-hoc comparisons were used to evaluate permeability reduction differences among groups (α = 0.05). Results: The null hypothesis was rejected (p 0.0001). The data provided strong evidence that the distribution of percentage permeability reduction differed among the 5 treatment groups (p 0.0001). All possible pairwise comparisons of the groups were statistically significant. The highest mean percentage of permeability reduction was observed in the group where manufacturer's instructions were followed (99.4 ± 1.0), followed by the aggressive drying primer group (86.7 ± 11.1), the shortened primer application group (73.01 ± 12.0), the no-drying primer group (48.44 ± 19.9), and, finally, the no-primer group (22.33 ± 7.7). Conclusion: The incorrect primer application allowed significant dentin permeability. Strict adherence to recommended clinical application of dentinal adhesives is fundamental to achieve good dentin tubule sealing. Keywords: dentin bonding, permeability