Purpose: Because the evidence regarding the efficacy of splinting short implants on strain distribution is limited, this study aimed to measure and compare the strains generated by splinted and nonsplinted screw-retained implant crowns for short implants with internal connections. Material and Methods: Computed tomographic scan data of a patient missing all mandibular molar teeth were digitized using a software program, and stereolithography was then used to produce an acrylic resin cast. Two 4- × 6-mm implants were placed in the left side. One set of splinted crowns and one set of nonsplinted crowns were made to fit the two implants using screw retention. Three-dimensional image correlation was used for full-field measurement of strains and provided a synchronized stereo view of the cast during the experiment. Cameras recorded changes in random dot patterns that had been applied to the surface of the cast as the prostheses were loaded up to 400 N in vertical and oblique directions using a universal testing machine. Testing was repeated three times for splinted and nonsplinted crowns. An image correlation algorithm used the dot pattern to define correlation areas. Coordinates of these areas were determined for each recorded photograph and used to calculate the strains. Strain distribution data were compared for maximum and minimum principal strains. Results: Qualitative data showed evidence of increased load sharing for splinted, screw-retained prostheses regardless of the direction of the load applied. However, the only statistically significant difference between the two prosthesis types occurred for peak maximum principal strain under oblique loading. Conclusions: Splinting short implants may provide a more even strain distribution during functional loading. Clinical corroboration of these findings is needed. Palabras clave: screw-retained, short implants, splinted