||Gradient elution theory can be used as a powerful tool for predicting retention and thereby optimizing separations in reversed-phase high performance liquid chromatography (HPLC). One of the objectives of this research was to extract isocratic parameters from gradient retention data, and to use this information to accurately predict optimum separation conditions. Ultimately, the goal was to apply this strategy to the separation of biological macromolecules, such as peptides and proteins, for which gradient elution becomes a necessity. However, the optimum use of gradient elution for method development depends on exact relationships between isocratic and gradient retention data. An evaluation of this correlation was first undertaken for small test molecules, a series of dialkyl phthalates. From this research, several factors were found which can influence solute retention in a gradient. Equipment design and operating conditions can distort the gradient profile, and a theoretical study shows how this affects solute retention times. Other 'non-ideal' processes are due to the column/mobile phase system. The most important of these are the nonlinear relationship of log k' and mobile phase composition, solvent demixing, which results from adsorption of one mobile phase component onto the column, and the variation in column dead-time, t(,o), with mobile phase composition and flow rate. This information can be used to correct experimental retention data or in choosing experimental conditions which minimize these non-ideal effects. After applying these corrections to experimental data, calculated values of gradient retention time, t(,g), were in agreement with experimental values within (+OR-)1% (1 std. dev.) of the total gradient time t(,G). There has been much controversy over the mechanism of the chromatographic process for macromolecules. The isocratic and gradient retention of polystyrene standards was studied, using a broad range in standard molecular weights. From this study, it was found that the retention model developed for small molecules can be applied to macromolecules with equal precision. Finally, gradient theory was used to optimize the separations of complex mixtures of biological macromolecules. By varying gradient parameters, such as flowrate and gradient time, a method was developed for predicting optimum separation conditions for these complex samples.