PURPOSE: The aims of this study were to evaluate the ability of a stepped osteotomy to improve dental implant primary stability in low-density bone sites and to investigate possible correlations between primary stability parameters. MATERIALS AND METHODS: The study was performed on fresh humid bovine bone classified as type III. The test group consisted of 30 Astra Tech EV implants inserted following the protocol provided by the manufacturer. The first control group consisted of 30 Astra Tech EV implants inserted in sites without the underpreparation of the apical portion. The second control group consisted of 30 Astra Tech TX implants inserted following the protocol provided by the manufacturer. Implant insertion was performed at the predetermined 30 rpm. The insertion torque data were recorded and exported as a curve; using a trapezoidal integration technique, the area underlying the curve was calculated: this area represents the variable torque work (VTW). Peak insertion torque (pIT) and resonance frequency analysis (RFA) were also recorded. RESULTS: A Mann-Whitney test showed that the mean VTW was significantly higher in the test group compared with the first control and second control groups; furthermore, statistical analysis showed that pIT also was significantly higher in the test group compared with the first and second control groups. Analyzing RFA values, only the difference between the test group and second control group showed statistical significance. Pearson correlation analysis showed a very strong positive correlation between pIT and VTW values in all groups; furthermore, it showed a positive correlation between pIT and RFA values and between VTW and RFA values only in the test group. CONCLUSION: Within the limitations of an in vitro study, the results show that stepped osteotomy can be a viable method to improve implant primary stability in low-density bone sites, and that, when a traditional osteotomy method is performed, RFA presents no correlation with pIT and VTW.