This paper investigates the application of X-ray micro-computed tomography (micro-CT) to accurately evaluate bone formation within 3D printed, porous Ti6Al4V implants manufactured using Electron Beam Melting (EBM), retrieved after six months of healing in sheep femur and tibia. All samples were scanned twice (i.e., before and after resin embedding), using fast, low-resolution scans (Skyscan 1172; Bruker micro-CT, Kontich, Belgium), and were analysed by 2D and 3D morphometry. The main questions posed were: (i) Can low resolution, fast scans provide morphometric data of bone formed inside (and around) metal implants with a complex, open-pore architecture?, (ii) Can micro-CT be used to accurately quantify both the bone area (BA) and bone-implant contact (BIC)?, (iii) What degree of error is introduced in the quantitative data by varying the threshold values?, and (iv) Does resin embedding influence the accuracy of the analysis? To validate the accuracy of micro-CT measurements, each data set was correlated with a corresponding centrally cut histological section. The results show that quantitative histomorphometry corresponds strongly with 3D measurements made by micro-CT, where a high correlation exists between the two techniques for bone area/volume measurements around and inside the porous network. On the contrary, the direct bone-implant contact is challenging to estimate accurately or reproducibly. Large errors may be introduced in micro-CT measurements when segmentation is performed without calibrating the data set against a corresponding histological section. Generally, the bone area measurement is strongly influenced by the lower threshold limit, while the upper threshold limit has little or no effect. Resin embedding does not compromise the accuracy of micro-CT measurements, although there is a change in the contrast distributions and optimisation of the threshold ranges is required.