Four types of implant superstructures were screwed onto implant bodies, and the strains created around the implant bodies were compared and analyzed within the IMZ Implant System. Three IMZ implants were embedded in the center of a polyurethane block (30 x 40 x 30 mm), and a total of 16 superstructures was fabricated by 4 methods: 1-piece cast, 1-piece cast/split soldering, soldering, and passive fit. Six strain gauges were placed on the surface of the block 1 mm apart. Three embedded implants were numbered, and a fixed partial denture was placed on these implants and screwed by a torque wrench using 14.5 Ncm torque. This procedure was repeated 7 times for each fixed partial denture, and each created strain was measured when the last screw was tightened. In all fixed partial dentures, strains were produced around the implant bodies when screws retaining the prosthesis were tightened, and the strain was relieved with unscrewing. The magnitude of strain was greater with the 1-piece cast method or the section/solder method than with the soldering and passive-fit methods. Of the 2 soldering methods, when the screw on the middle implant was tightened before those on the terminal 2 implants, the magnitude of strain was lower with the soldering method than with the 1-piece cast/split soldering method. When the order of screw tightening was changed, there were significant differences in the magnitude of strain at each gauge with the soldering method. With the passive-fit method, no differences in the magnitude of strain attributable to the order of screw tightening could be detected. The magnitude of strain produced around a screw-retained implant prosthesis was significantly lower with the passive-fit method when compared to the other 3 fabricating methods. Furthermore, the implants prepared by the passive-fit method were not affected by the order of screw tightening. Keywords: implant prosthesis, passive fit, strain gauge, stress distribution, tightening screw