PURPOSE. This study evaluated the initial stability of different implants placed above the bone level in different types of bone. MATERIALS AND METHODS. As described by Lekholm and Zarb, cortical layers of bovine bone specimens were trimmed to a thickness of 2 mm, 1 mm or totally removed to reproduce bone types II, III, and IV respectively. Three Implant system (Br${aa}$nemark System$^{(R)}$ Mk III TiUnite$^{TM}$, Straumann Standard Implant SLA$^{(R)}$, and Astra Tech Microthread$^{TM}$-OsseoSpeed$^{TM}$) were tested. Control group implants were placed in level with the bone, while test group implants were placed 1, 2, 3, and 4 mm above the bone level. Initial stability was evaluated by resonance frequency analysis. Data was statistically analyzed by one-way analysis of variance in confidence level of 95%. The effective implant length and the Implant Stability Quotient (ISQ) were compared using simple linear regression analysis. RESULTS. In the control group, there was a significant difference in the ISQ values of the 3 implants in bone types III and IV (P<.05). The ISQ values of each implant decreased with increased effective implant length in all types of bone. In type II bone, the decrease in ISQ value per 1-mm increase in effective implant length of the Br${aa}$nemark and Astra implants was less than that of the Straumann implant. In bone types III and IV, this value in the Astra implant was less than that in the other 2 implants. CONCLUSION. The initial stability was much affected by the implant design in bone types III, IV and the implant design such as the short pitch interval was beneficial to the initial stability of implants placed above the bone level.