Since the concept of osseointegration was introduced by Branemark of Sweden, dental implants have been used for various dental prosthetic treatments. The survival rate of dental implant is likely to be closely related to the total biomechanical role of each component of implant system. The use of interchangeable component is very attractive for dental practitioners because such an approach would save treatment cost, flexibility of prosthetic treatment options as well as conveniences. Therefore, the use of interchangeable implant system has been increasing without scientific assessment of safety and efficacy of various interchangeable implant system. The purpose of this study, therefore, were to compare the geometric characteristic of four interchangeable dental implant screws and the loosening torque of these screws. Four types of dental implant screws tested in this study were Nobelpharma, 3i, Impla-med, Restore. Four screws each of the test specimens were subjected for scanning electron microscopic examinations under the same condition and a 35x magnified standard SEM picture was objected from each test specimen using JSM-5200 scanning microscope. From each of the SEM pictures, eight parameters. i.e., diameter of screw head. screw length, thread pitch, major diameter. neck diameter, neck length, crest width and root width were determined using a caliper. The measurement for each parameters were then corrected for their magnification factor. The loosening torque were also determined by using a torque gauge. All of the measurements were statistically analyzed by ANOVA test and multiple range test. Statistical significance was set in advance at the probability level of less than 0.05. All analyses were done with SPSS software for the personal computer. The conclusion obtained from this studies were summarized as the following; 1. No statistically significances were noted in the thread pitch. and crest width in the four screws, and in the case of major diameter, the Impla-med screw was significantly smaller than the other three screws (p<0.05). Therefore, four implant bolts could be physically inserted in a abutment nuts. 2. The diameter of screw head was decreased in the order of Restore, 3i, Nobelpharma, Impla-med screws and the length of screws were decreased in the order of 3i, Restore, Nobelpharma and Impla-med. The diameter of neck was decreased in the order of Impla-med, Restore. Nobelpharma, 3i screws. The differences of each of these parameters were statistically significant (p<0.05). The width of root of screws were decreased in the order of Nobelpharma, Impla-med, Restore and 3i. The differences among Nobelpharma and Impla-med. Restore and 3i were statistically significant (p<0.05). 3. When the screws were loosening 1, 3, 4 and 5 times, the loosening torque for Impla-med and 3i screws were significantly higher than that of Nobelpharma or Restore screws (p<0.05). However, when statistically smaller than that of 3i, Restore or Nobelpharma screws(p<0.05).