Bonding of composite resins to sound and caries-affected dentin in cervical areas may necessitate the use of hemostatic agents to control sulcular fluid and hemorrhage. The aim of this in vitro study was to evaluate the bond strengths of a self-etching adhesive system to sound and caries-affected dentin after the use of 3 different hemostatic agents. Composite resin cylinders were bonded to 48 caries-affected and 48 sound dentin surfaces in 8 groups. Groups 1-4 utilized caries-affected dentin: group 1, uncontaminated control; 2, ViscoStat; 3, ViscoStat Clear; and 4, trichloroacetic acid (TCA). Groups 5-8 utilized sound dentin: group 5, uncontaminated control; 6, ViscoStat; 7, ViscoStat Clear; and 8, TCA. The hemostatic agents were applied for 2 minutes and rinsed. After 500 rounds of thermocycling, shear bond strength tests were carried out. Data were analyzed with 1- and 2-way analyses of variance, t test, and post hoc Tukey tests at a significance level of P < 0.05. Bond strength was significantly influenced by dentin type (F = 38.23; P = 0.0001) and hemostatic agent (F = 6.32; P = 0.001). Furthermore, groups 2 and 6 (ViscoStat) showed significantly lower bond strength values than the control groups (groups 1 and 5) in both affected and sound dentin (P = 0.043 and P = 0.009, respectively). Within the limitations of this study, the bond strength of composite resin to caries-affected dentin was significantly reduced compared to that with sound dentin. Among the studied hemostatic agents, ViscoStat resulted in a greater decrease in dentin bond strength. Contamination of both sound and caries-affected dentin with hemostatic agents decreased composite resin bond strength. Of the 3 hemostatic agents used, ViscoStat Clear appeared to have the least detrimental effect on bond strength.