We used our established experimental model of revision joint replacement to examine the roles of hydroxyapatite coating and bone graft in improving the fixation of revision implants. The revision protocol uses the Søballe micromotion device in a preliminary 8-week period of implant instability for the presence of particulate polyethylene. During this procedure, a sclerotic endosteal bone rim forms, and a dense fibrous membrane is engendered, having macrophages with ingested polyethylene and high levels of inflammatory cytokines. At the time of revision after 8 weeks, the cavity is revised with either a titanium alloy (Ti) or a hydroxyapatite (HA) 6.0 mm plasma-sprayed implant, in the presence or absence of allograft packed into the initial 0.75 mm peri-implant gap. The contralateral limb is subjected to primary surgery with the same implant configuration, and serves as control. 8 implants were included in each of the 8 treatment groups (total 64 implants in 32 dogs). The observation period was 4 weeks after revision. Outcome measures are based on histomorphometry and mechanical pushout properties. The revision setting was always inferior to its primary counterpart. Bone graft improved the revision fixation in all treatment groups, as also did the HA coating. The sole exception was revision-grafted HA implants, which reached the same fixation as primary Ti and HA grafted implants. The revision, which was less active in general, seems to need the dual stimulation of bone graft and HA implant surface, to obtain the same level of fixation associated with primary implants. Our findings suggest that the combination of HA implant and bone graft may be of benefit in the clinical revision implant setting.