Three types of calcium phosphate coating were formed on polyethersulphone (PES) rectangular plates using a biomimetic method: a 20 microns thick amorphous calcium phosphate (ACP 20) coating, a 50 microns thick amorphous calcium phosphate (ACP 50) coating, and a 50 microns thick highly crystalline hydroxyapatite (hHA 50) coating. Uncoated PES plates were used as a control group. These materials were implanted in the tibiae of rabbits and subcutaneously in rats, and the samples were harvested 8 and 16 weeks thereafter, and were examined histologically. The tensile failure loads at the bone-implant interfaces were determined using the detaching test. Each ACP coating was more degradable than the hHA 50 coating. However, newly formed bone came into direct contact with underlying materials as the coating degraded. No coating degraded in subcutaneous tissue. Soft tissue intervening was seen in uncoated samples. Failure load of ACP 20-, ACP 50- and hHA 50-coated samples were all relatively higher than that of the uncoated samples at each period. Significant increase of failure load was seen in hHA 50-coated samples by 16 weeks, however, no increase was seen in either the uncoated or ACP-coated samples. If coating longevity is desired, then the hHA coating is preferable. However, if only the osteoconducive property of calcium phosphate coating is desired for initial fixation of porous materials, the ACP coating may be advantageous.