In the present study, we tested the in vitro process of differentiation and mineralization as well as the process of in vivo bone formation on substrates with different macrostructures. We used carbonated apatite-coated titanium discs that were respectively smooth, plasma spayed with titanium or had a porous structure. Subcultured rat bone marrow cells were seeded on the substrates and after 7 days of culture, the tissue-coated substrates were subcutaneously implanted in nude mice for 4 weeks. After 1 week of culture in the presence of the osteogenic differentiation promoter dexamethasone, the cells had formed a continuous layer of mineralized tissue on the smooth and titanium plasma-sprayed discs. In the case of the porous titanium discs, the bone-like tissue coverage was restricted to the outer surface and the peripheral pores. The influence of the macrostructure on the process of differentiation of the cultured cells depended on the presence of dexamethasone. When dexamethasone was present, the highest ALP/DNA ratios were obtained with the smooth surfaces. In the absence of dexamethasone, the highest ALP/DNA values were obtained with the rough macrostructured discs. We postulate that these different patterns were due to the shielding of cells in pits or pores of rough structured substrates by dense overlying cell layers. These cell layers are suggested to increase the exposure of excreted osteoinductive proteins and decrease the exposure of dexamethasone to underlying cells. Four weeks post-implantation, abundant bone formation could be observed on all in vitro tissue-coated substrates. The percentage of direct bone contact on the porous discs (42.3 +/- 22.3) was significantly lower compared to the non-porous discs. This was related to the process of bone infiltration into the central oriented pores that predominantly occurred in a centrifugal manner. The percentage of direct bone contact on the smooth discs (96.3 +/- 2.3) was significantly higher compared to the titanium plasma-sprayed discs (81.5 +/- 10.7). This was not due to fibrous tissue infiltration, but due to the extensive formation of bone marrow. Nevertheless, for practical reasons regarding protection of the layer of cultured cells during the implantation procedure, the use of rough or porous surface structures is suspected to be advantageous in revision surgery.