Implant stability is the key to long-term successful outcome for osseointegrated implants. To evaluate the initial healing response of bone around HA-coated implants without primary bone contact. 21 HA-coated thread type implants(STERI-OSS?) were placed in the femurs of 5 mongrel dogs, about 1-year old. Implants, 8 mm in length and 3.8mm(experimental 1group), 5.0mm(experimental 2group) and 6.0mm(control group) in diameter, were inserted after 3 holes of 6.0mm in diameter and 10mm in depth were prepared in the surgical sites each dog. Implants were supported by only nonresorbable membrane($Teflon^{(R)}$), in order to prevent the ingrowth of upper soft tissue into the gap between bone and implant, and to maintain each implant to be positioned in the center of the drilled hole. 9 implants with different diameters were inserted in 3 dogs for histologic observation, and 12 implants were inserted in 2 dogs for mobility test and removal torque test. Fluorescent dyes were injected for the observation of new bone formation in order of $Terramycin^{(R)}$, Arizarin $Red^{(R)}$, and $Calcein^{(R)}$ at an interval of 2 weeks. 3 dogs were sacrificed for histologic observation at 4, 8, and 12-week after placement. Light microscopy and confocal laser scanning microscopy were used to qualitatively characterize the bone around HA-coated implant. 2 dogs were sacrificed for mobility test($Periotest^{(R)}$, Simens AG, Bensheim, Germany) and removal torque test($Autograph^{(R)}$ AGS-1000D series, Japan) at 8 and 12-week after placement The results were as follows: 1. Histologic observation showed that osseointegration occurred to both control and experimental groups as time lapse, but delayed bone healing was revealed in 3.8mm group (experimental 1group), compared to contrtol group and 5.0mm group (experimental 2group). 2. The mobility test showed that the experimental groups had no distinguishable movement during experimental periods of 8 and 12-week, and there was no difference in mobility depending on the gap between bone and implant, and time lapse. 3. The removal torque forces were increased depended on the gaps decreasing between bone and implant, and time lapse. The results suggest that HA-coated implant without primary bone contact, based on guided bone regeneration could obtain its stability in all experimental groups as time lapse, but bone healing was delayed in experimental group of 3.8mm. And the results suggested that studies on correlationship between mobility test and removal torque test for implant stability would be necessary.