Aim: The aim of this in vitro study was to determine the influence of different techniques of retrograde cavity preparation on bond strength of different retrograde filling materials. Materials and methods: In this study, 138 extracted, human, single-rooted teeth were used. Teeth crowns were sectioned, followed by root canal instrumentation with rotating ProTaper instruments (Dentsply Maillefer, Ballaigues, Switzerland) and obturation with ProTaper F3 gutta-percha points and AH Plus endodontic sealer (Dentsply, DeTrey, Konstanz, Germany). Roots were sectioned 3 mm short of the apex and randomly divided into four groups (n=32). Retrograde cavities, 5 mm deep, were prepared with an ultrasound device in the first experimental group, in the second with Er:YAG laser, in the third with Er,Cr:YSGG laser, in the fourth using a round steel bur. Each group was subdivided into three subgroups (n=10), depending on the retrograde filling material used (MM-MTA (MicroMega, Besançon Cedex,France)), Biodentine (Septodont, Saint Maur-des Fosses, France), or TotalFill RRM (FKG Dentaire, La Chaux-de-Fonds, Switzerland) to fill the retrograde cavities. A control subgroup consisted of retrograde cavities prepared with a round steel bur and filled with amalgam (n=10). The roots were stored at 37 °C in a 100% moist environment. After one week, the roots were embedded in acrylic resin (Denta-O-Resin, M+W Dental, Büdingen, Germany). The apical part of each root was cut perpendicular to the long axis into 1 mm thick slices with a diamond blade and a push-out test was performed using a universal testing machine (double-column 3300 series, Instron, Illinois, USA). Bond strength (in MPa) was calculated. Failures were divided into adhesive, cohesive or mixed after observation with the stereomicroscope (06x-50x). Two roots from each group, with retrograde cavities, but without retrograde filling material placed, were examined under a scanning electron microscope (SEM) (JSM-6060LV, JEOL, Tokyo, Japan). For statistical analysis Kolmogorov-Smirnov test, one- and two-way ANOVA and χ2 tests were used with a level of significance set at 5%. Results: The two-way ANOVA showed that retrograde filling material factor had a significant influence (p=0.008) on the bond strength values, as well as the factor of retrograde cavity preparation technique (p=0.004). A significant interaction between the considered factors was found (p=0.001) meaning that the different retrograde materials tested showed different bond strength values depending on the type of retrograde cavity preparation. One- way ANOVA showed that the factor retrograde filling material influenced the bond strength when retrograde cavities were prepared by ultrasound or steel bur, while there were no significant differences among tested materials when both lasers were used. Among the group of retrograde cavities made by ultrasound, material TotalFill RRM had significantly higher bond strength than MM-MTA (p=0.010). Among the group of retrograde cavities made by a steel bur, the material MM-MTA had a significantly higher bond strength value compared to TotalFill RRM (p=0.011), Biodentine (p=0.003) and amalgam (p=0.002) (the control subgroup). The technique of retrograde cavity preparation factor influenced the bond strength in case of Biodentine (p=0.048) and TotalFill RRM (p=0.001) materials. In the case of Biodentine, the ultrasound technique showed a significantly higher bond strength result compared to the steel bur technique (p=0.042), while in the case of TotalFill RRM material the ultrasound technique showed a significantly higher bond strength value compared to the steel bur (p=0.001) and Er,Cr:YSGG laser (p=0.045) techniques, and Er:YAG laser technique significantly higher than the steel bur (p=0.002) technique. The highest mean bond strength was achieved in the group ultrasound retrograde cavity preparation technique (10.95 MPa). There were significant differences in mean bond strength values among the tested cavity preparation techniques. The material TotalFill RRM had the highest mean bond strength value (10.23 MPa) of all the materials. There were significant differences in mean bond strength values among the tested materials. The highest single bond strength value was recorded in the ultrasound technique/TotalFill RRM material subgroup (13.04 MPa). The most common type of failure was the mixed type, in all subgroups except in Er:YAG/TotalFill RRM and bur/MM-MTA subgroups where the adhesive failure type prevailed and the cohesive type of failure was not found. The second most common type of failure was adhesive and the cohesive type was the rarest. Conclusion: The second null hypothesis could be rejected, since according to the results of this study the TotalFill RRM filling material had a better bond strength to retrograde cavities prepared with ultrasound tip, and MM-MTA had a better bond strength to drilled retrograde cavities than the other material-cavity combinations tested. It was also found that the bond strength of the tested materials was highly dependent on the type of cavity preparation. The first null hypothesis could, therefore, also be rejected. These results indicate that the best material or the best cavity preparation technique among the tested ones could not be found, but a chosen association between a material and a cavity preparation technique could obtain excellent bond strength results, which could be almost three times higher than other associations.