Purpose: The aim of this study was to compare the bone stress generated by rehabilitation using regular and short-length Morse taper implants (11 and 5 mm, respectively) in the same context and allied with splinted (SP) and nonsplinted (NSP) prostheses in the posterior area of the mandible using finite element analysis. Materials and Methods: Three-dimensional geometric models using regular implants (∅4 × 11 mm) and short implants (∅4 × 5 mm) were simulated with a left posterior mandible that had the first premolar tooth and all teeth posterior to that premolar removed. The four experimental groups were as follows: Group 1 NSP (two regular implants and one short implant rehabilitated with nonsplinted prostheses), Group 1 SP (two regular implants and one short implant rehabilitated with splinted prostheses), Group 2 NSP (one regular implant and two short implants rehabilitated with nonsplinted prostheses), and Group 2 SP (one regular implant and two short implants rehabilitated with splinted prostheses). Oblique forces were simulated in molars (365 N) and premolars (200 N). Qualitative and quantitative analyses of the minimum principal stress in bone were performed using the ANSYS Workbench software, version 10.0. Results: The splinted prostheses decreased the stress to the surrounding bone of short implants. Moreover, they also decreased the stress on the surrounding bone adjacent to the tooth. However, the splinted prostheses generated higher stress in the coverage of the surrounding bone for regular intermediate implants and prominently in the bone at the implant cervical region, at the edge crest, compared with the individualized groups. Conclusion: The benefits in the use of splinted prostheses are notable for the preservation of the bone surrounding the short implants and tooth; however, it is necessary to evaluate each clinical situation because, in this context, the regular implants are at higher stress than the individualized implants. Parole chiave: ceramics, dental implantation, dental prosthesis implant-supported, finite element analysis