The subjects of investigation were two submicron powders with grain sizes of 0.75 μm and 0.6 μm and a cobalt contents of 10 % and 8.2 % respectively. The pow- ders were compacted using direct, dry-bag and extrusion moulding. Samples were drawn after presintering the materials to represent the raw moulded material, after sintering, after sinter HIP and after treating the sintered samples with HIP. In order to determine which material, moulding process and temperature-pressure treatment combination would lead to the best result, all samples were metallographically in- vestigated, especially evaluating the resulting porosity. The presintered state showed a clear advantage in microstructure for the samples, made from powder with smaller grainsize and less cobalt content. These samples also show a higher porosity after sintering, manufactured by extrusion and dry-bag moulding. The reason for this is the smaller grainsize of the material, which equals a higher specific surface and narrower grainsize distribution, which results in the grains not packing well together and more small pores between the grains. Due to this, more rearrangement of the material structure and more shrinkage during the sintering process is necessary, in order to fill the voids. This can lead to a higher porosity especially if the structure rearranges itself aiming to reduce the specific surface. Extrusion moulded samples of either material show higher porosity than the the samples of other manufacturing methods. Surprising is the higher amount of B-pores in extrusion moulded HIP samples than sinter HIP samples, even though HIP operates with a much higher pressure. Dry-bag moulding seems to be the best manufacturing method for both materi- als, closely followed by direct moulding. Sinter HIP doesn’t seem to cause any kind of disadvantage concerning the porosity in either of the tested materials, despite the lower pressure, compared to HIP. Concerning the automatic investigation of porosity via optical light microscope it is likely that the process of detecting pores is too prone to falsification due to dirt, alcohol staines, dust, etc. and also too complex to be reduced to different shades of grey.