Purpose: The aim of this study was to compare the biomechanical performance of splinted or nonsplinted prostheses over short- or regular-length Morse taper implants (5 mm and 11 mm, respectively) in the posterior area of the mandible using finite element analysis. Materials and Methods: Three-dimensional geometric models of regular implants (Ø 4 × 11 mm) and short implants (Ø 4 × 5 mm) were placed into a simulated model of the left posterior mandible that included the first premolar tooth; all teeth posterior to this tooth had been removed. The four experimental groups were as follows: regular group SP (three regular implants were rehabilitated with splinted prostheses), regular group NSP (three regular implants were rehabilitated with nonsplinted prostheses), short group SP (three short implants were rehabilitated with splinted prostheses), and short group NSP (three short implants were rehabilitated with nonsplinted 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 ANSYS Workbench software, version 10.0. Results: The use of splinting in the short group reduced the stress to the bone surrounding the implants and tooth. The use of NSP or SP in the regular group resulted in similar stresses. Conclusions: The best indication when there are short implants is to use SP. Use of NSP is feasible only when regular implants are present. Keywords: ceramics, dental implantation, dental prosthesis implant-supported, finite element analysis