Graphene, a honeycomb lattice of a single atom thick plane of carbon atoms has captivated the attention of physicists, materials scientists, and engineers because of its extraordinary physical properties. These include exceedingly high charge carrier mobility, current-carrying capacity, thermal conductivity, and mechanical strength. These properties qualify it as an excellent candidate for new electronic technologies both within and beyond metal oxide semiconductors. The attractive properties of graphene for future applications are observed only in high quality graphene produced by mechanical exfoliation of graphite by using the so-called "scotch tape method". However, the yield is low and it satisfies only the needs of basic research. For applications one has to have material which is produced on large scale but preserves the quality of graphene. There are many production routes proposed in the scientific community and their number is still growing. One of the tasks of this PhD work was to test by Electron Spin Resonance (ESR) the quality of graphene produced by various methods. ESR is a suitable technique for looking for ferromagnetic interactions in graphene predicted by theory, and to determine the intrinsic spin relaxation time? These features are important for integration of graphene in spintronics applications. The main results are obtained on a large assembly of three differently prepared samples: 1. Mechanically exfoliated graphite (MEG), 2. Reduced graphene oxide (RGO) and 3. Liquid phase exfoliated (LPE) graphite. Attempts were done to study graphene synthesized by chemical vapor deposition (CVD) and by epitaxial growth on SiC. The mother compound graphite, as a reference sample was also thoroughly investigated by ESR. The LPG sample, obtained by heavy sonication of graphitic powder in NMP solvent, contains more nanographitic particles than graphene. The overall response resembles that of graphite with the exception of detecting a strong magnetic interaction below 26 K. This unambiguously shows the possibility of creating a strong magnetic response in carbon based material, but the challenge to inducing it graphene still remains. The RGO sample, prepared by chemical methods provides a large quantity of sample, suffers from the harsh oxidative and reduction steps involved in the synthesis. Both steps can be incomplete, leaving behind a large number of defects which Anderson-localize the extended states necessary for electronic applications. The best characteristics are obtained on MEG graphene. The temperature dependence of the magnetic susceptibility follows that predicted by theory, and it is distinctly different from that of graphite. Furthermore, the 10-8 s spin lifetime deduced from the ESR linewidth is by two orders of magnitude longer than that obtained in spin-valve experiments and it is encouraging for spintronics applications.