The formulation of the Theory of General Relativity and the observational evidence for the expansion of the universe provided the basis for much of the work carried out in the field of cosmology over the past hundred years. Huge volumes of research have been conducted to find reliable values for cosmological parameters and to describe the amount and nature of the matter in the universe. Chapter 1 of this thesis attempts to summarise current theoretical and observational thinking on these matters and, in particular, examines the wide-ranging application of gravitational lensing to the search for so-called dark matter. The use of gravitational microlensing to investigate a cosmological population of compact objects, their effects on the long term variability of the apparent luminosity of quasars and on the results of the on-going observations of high redshift supernovae is discussed. Such investigation forms the basis for this thesis. The main tool for this investigation is a computer model which simulates the gravitational lensing effect of a population of compact object over a period of time. Chapter 2 sets out the theoretical background for this simulation. In particular, the methods used to set the physical parameters of the simulation, such as its volume, the redshifts of the lenses and their masses, are outlined. Chapter 3 presents the implementation of the computer model. Modelling techniques used by other researchers are discussed, as are alternative approaches considered for the implementation of this model. In order to simulate the evolving distribution of the lensing objects over time, the simulation was designed to run on high performance parallel supercomputers. The method by which the simulation was designed to take advantage of this type of computing platform is also discussed. In order to examine the effects of a cosmological distribution of compact objects on high redshift sources properly, it is necessary to have observational data. For this thesis, the observational data consists of a set of lightcurves from high redshift quasars observed over a 25 year period. This data set is outlined in Chapter 4. The results from the computer simulation are then presented, including both example light curves and power spectra for a variety of cosmological models, source sizes, source redshifts and lens masses. This observational data is compared with the simulation data and is found to have comparable levels of power for a number of simulation models. Chapter 5 examines the effect of a cosmological population of compact objects on the ongoing high redshift supernovae searches. The effects of such objects are modelled for a number of cosmological models for the range of redshifts proposed for the SNAP and VISTA searches. It is found that the proposed number counts for supernovae detection in each redshift bin are sufficient to differentiate between the different cosmological models.