In this study, we report an experimental attempt to resolve whether ferromagnetism of graphite nanoflakes is an intrinsic phenomenon or it solely originates from impurities. A comparative study of either a nominally undoped or intentionally contaminated with NiO or Gd2O3 samples was performed. We show, first, that a detectable by X-ray diffraction contamination may occur via the agate mortar/pestle working surfaces if prior to sample dispersion it was used for grinding of hard oxides. Second, we find a systematic trend in a development of a FM component of all three samples under vacuum annealing at 400 or 800◦C. Third, we notice that the samples notably contaminated with NiO or Gd2O3 do not reveal any drastic enhancement in ferromagnetism with respect to the sample free from intentional doping, contrary to an expectation related to nickel and gadolinium oxides reduction to metallic ferromagnetic at room temperature state. As a result, we conclude that ferromagnetism of graphite nanoflakes is probably an intrinsic phenomenon that could be stimulated slightly by NiO or Gd2O3 impurities, though an impact of the agate (SiO2) contamination itself may also play a role.