Multiple collector inductively coupled plasma mass spectrometry (MC-ICPMS) has been used for the measurement of Fe and Zn isotope ratios in a variety of sample types including natural and synthetic waters, ferromanganese concretions, sediments and biological material. The focus of this PhD study was the development of analytical procedures for precise and accurate measurement of Fe and Zn isotope ratios by high resolution MC-ICPMS. The objective of the study has also been to complement the existing knowledge of Fe and Zn isotope fractionation that occurs due to fundamental physico- chemical processes such as oxidation-reduction and diffusion. Prior to the MC-ICPMS measurements, anion-exchange chromatography and precipitation procedures for purification of Fe and Zn have been thoroughly evaluated in terms of recoveries, decreases in matrix elements concentrations and elimination of interfering species. To correct for differences in instrumental mass discrimination effects between samples and standards, on-line normalisation procedure has been employed using Ni and Cu as elemental spikes for Fe and Zn isotope ratio measurements, respectively. It has been demonstrated that the on-line normalization in combination with sample-standard bracketing technique adequately correct for drift in the instrumental mass discrimination and reduce standard deviation of Fe and Zn isotope ratio measurements. External reproducibilities of 56Fe/54Fe and 66Zn/64Zn isotope ratios at typical concentrations of 2-5 mg l-1 were better than 50 ppm and 30 ppm (one standard deviation), respectively. Based on the established analytical methods Fe isotope variations have been explored at the oxic-anoxic interface of seasonally anoxic lake water and in deposited sediments under different redox conditions. The observed distribution of d56Fe values during redox cycling of Fe was found to be consistent with the model of equilibrium Fe isotope fractionation between co-existing aqueous Fe (II) species, such as Fe(OH)+, prior to oxidation. Different patterns of d56Fe values distribution for Fe oxide phase have been identified in the deposited sediments. In anoxic sediment column, where anoxic conditions were unchanged during the time of deposition, the d56 values are near-zero and rather invariant. In contrast, in sediment column of permanently oxygenated lake, where zone of the oxic-anoxic transition moved over time, d56 values for Fe oxides show large shifts, spanning the range of ~ 1.2‰. In light of the use of Fe isotope data in inference of geochemical pathways for Fe that may have occurred during primary precipitation and diagenesis, these data show the isotopic signature of Fe oxide phase that was accumulated due to precipitation of oxidized ferrous Fe flux. The measurements of Fe and Zn isotope ratios in aqueous solutions sampled at varying distances from sources of these ions revealed fractionation of the isotopes resulting from pure diffusion in solution. It has been demonstrated that diffusion alone can cause changes in 56Fe/54Fe and 66Zn/64Zn isotope ratio in excess of ~ 0.3‰. Furthermore, the effect of diffusion in hydrogels of diffusive gradients in thin films (DGT) samplers for measurements of isotopic composition of soluble metals as a possible source of fractionation has been investigated by using Zn as a test element. It has been shown that, provided quantitative elution is obtained, no fractionation of Zn isotopes due to the diffusion process is detectable within the reported precision of MC-ICPMS measurements. Consequently, DGT samplers are suitable for studies of the isotopic composition of soluble metals that are not affected by redox processes.