To evaluate the effects of different hemostatic agents on the shear bond strength (SBS) in vitro and clinical bond failure rate of orthodontic metal brackets in vivo. A total of 100 human premolar teeth were randomly divided into five groups: control, blood, Viscostat, hydrogen peroxide (H2O2), and epinephrine. Teeth were bonded with same light-cured adhesive and composite. After storage in distilled water for 24h, thermal cycling was used as an aging procedure on all samples. The brackets were subjected to an SBS test at a speed of 0.5mm/min until bracket debonding. SBS values and the adhesive remnant index were evaluated. Ninety-nine patients (52 female, 47 males) undergoing routine orthodontic treatment were recruited for this controlled clinical study at bonding stages. All patients with bleeding on the buccal surface of any premolar tooth or teeth at bonding were included in this study. Over 6months, the bond failure rate was calculated. Data were analyzed using one-way analysis of variance (ANOVA) and Tukey's post-hoc test (p<.05). The McNemar test was used to compare bracket-bond failure. ANOVA showed a significant difference (p<.001) between the groups. No significant differences were found between the hemostatic agent groups (p>0.05) in the in vitro part. The lowest failure rate was obtained in the control group rather than the hemostatic agent groups during clinical follow-up (p<0.05). Each of the hemostatic agents (Viscostat, H2O2, and epinephrine) can be used for bleeding management during the orthodontic bonding process. Epinephrine application showed a high bond-failure rate at clinical follow-up.