Carbon nanofibers (CNFs) and carbon nanotubes (CNTs) have attracted intense research efforts with the expectation that these materials may have many unique properties and potential applications. The most promising way for large-scale synthesis of CNFs and CNTs is chemical vapor deposition (CVD). CNFs were synthesized on a series of hydrotalcite (HT) derived 77 wt.% Ni-Fe/Al2O3 catalysts in order to achieve the optimization of productivity and quality. It was found that only the Fe catalyst was active in CO disproportionation and only the Ni catalyst was active in ethylene decomposition, whereas all catalysts were active in ethylene decomposition when the reactants were a mixture of C2H4/CO. More control over the structure and diameter of the CNFs has been realized with the HT catalysts. At the same time, a high yield can be obtained. The synthesis process has been further studied as a function of various process parameters. It turned out that high hydrogen concentration, space velocity, and reaction temperature would enhance the production of CNFs. However, a slightly lower quality was associated with the higher productivity. The optimum CNF yield of 128 gCNF/gcat could be reached within 8 h on the HT catalyst with a Ni/Fe ratio of 6:1. Therefore, HT derived catalysts present a new promising route to large-scale controlled synthesis of CNFs. CNTs has been synthesized from CO disproportionation on Ni-Fe/Al2O3 supported catalysts with metal loadings of 20 and 40 wt.%. A high space velocity resulted in a high production rate but a short lifetime and a low carbon capacity. Increasing the metal loading to 40 wt.% significantly increased the reaction rate and productivity, and produced similarly uniform CNTs. Furthermore, H2 was found to be necessary for a high productivity, and the H2 partial pressure could be changed to adjust the orientation angle of the graphite sheets. The effects of catalyst particle size and catalyst support on the CNT growth rate during CO disproportionation were studied over SiO2 and Al2O3 supported Fe catalysts with varying particle sizes. It was found that there was an optimum particle size at around 13-15 nm for the maximum growth rate, and the growth rate was influenced both by the particle size and the support but the particle size was the dominating factor. The trends have been demonstrated at two different synthesis temperatures of 600 and 650°C. The effect of gas precursors on the yield and structure of carbon growth has been systematically investigated over powder Fe and Fe/Al2O3 catalysts. CO/H2, CO, CH4, and C2H6/H2 were the gas precursors studied. The carbon yield was higher on powder Fe from CO, but the yield was higher on Fe/Al2O3 from hydrocarbons. Completely different or similar carbon nanostructures were synthesized, depending on the gas precursors. It was suggested that the reactivity of gas precursors and the structures of carbon deposits are determined by the size and crystallographic faces of the catalyst particles, which are dictated by the interactions among metal particles, support, and the reactants. Controlled synthesis of CNT, platelet nanofiber, fishbone-tubular nanofiber, and onion-like carbon with high selectivity and yield was realized. A mechanism was proposed to illustrate the growth of different carbon nanostructures.