Many variables are involved in hydroxyapatite coating of metals by plasma-spray techniques. The authors have investigated the biological response to some of the most relevant variables in a controlled in vivo trial. The bone response in the rabbit towards hydroxyapatite coated cylinders was studied keeping the following variables fixed: (a) crystallinity of coating (greater than 90% and between 70% and 60%); (b) thickness of coating (50 and 100 mu m); (c) metallic substrate (titanium alloy and duplex stainless steel). Analysis of the results highlights the importance of defining the crystallinity of the coating to forecast its in vivo behaviour: highly crystalline coating is more stable in time but can give rise to fragmented bulky particles; a less crystalline coating is subject to slow degradation in the long term but facilitates its substitution by newly formed bone. Furthermore, it has been found that no relevant differences can be ascribed to a variation in coating thickness between 50 and 100 mu m. It has, also, been observed that there are no differences when duplex stainless steel is used instead of titanium alloy as metallic substrate, confirming that bone responds primarily to the coating.