In this thesis, I have developed a new methodology for the accurate determination of mass-dependent variations in nickel (Ni) isotope compositions. Nickel is initially separated in a three-column ion-exchange procedure, and the purified solutions are analysed by multi-collector inductively coupled plasma mass spectrometry (MCICPMS) using a double-spike technique. Using this methodology, I have measured the first Ni isotope ratios for a wide variety of natural geological samples. Significant Ni isotope variations were observed, with an overall spread in delta 60Ni-values of -0.9 to 2.5 permil. In igneous rocks Ni isotopes appear to be largely homogeneous, with only small variations (0.2 permil) between different rock types. Weathering of silicate rocks does on the other hand appear to cause significant fractionation of Ni isotopes, probably producing an isotopically heavy riverine input to the ocean. A heavy isotope signature is also visible in hydrogenetic ferromanganese crusts, with surface scrapings from globally distributed crusts show an average delta 60Ni-value of 1.65 permil. However, the variation in these samples is over 1.5 permil, likely reflecting local sources or biological processes, or alternatively indicating a heterogeneous Ni isotopic composition of the ocean. Organic-rich sediments also show heavy isotopic compositions, which are possibly transferred to the crude oils originating in these types of sediments. The only significant reservoir of light Ni isotopes found during this project are sulphides from magmatic systems. Overall, this thesis demonstrates the potential of this system as a powerful new tracer for a variety of geochemical processes.