Accurately predicting transport processes, including sediment transport, in the coastal environment is impossible without correct current velocity and shear stress information. A combined wave-current boundary layer theory is necessary to predict these quantities, and while the previous Grant-Madsen type boundary layer models are effective, they inconsistently apply a discontinuous two layer eddy viscosity structure to the wave and current problems. We have therefore developed a new continuous three layer model which consistently applies all three layers and leads to a strong coupling between the wave and current solutions. Boundary layer models require an estimate of the movable bed roughness, and while this roughness is scaled by the sand grain diameter for flat beds, in the coastal environment it is often the case that either wave-generated ripples cover the bed or the near-bed sediment is transported as sheet flow, in which case the roughness is much larger and less straightforward to characterize. The common method of predicting roughness in the ripple regime, while effective, unnecessarily predicts ripple geometry and requires a model-dependent factor, which varies widely, relating ripple geometry and bottom roughness. We have therefore developed an alternative, more direct method of predicting bed roughness: the wave energy dissipation factor is predicted from flow and sediment information and then any desired theoretical friction factor model is used to back-calculate the roughness. This proposed method can also be used in the sheet flow regime, allowing a continuous transition between the two regimes, not possible with the common method. This thesis derives the new three layer combined wave-current boundary layer theory, develops the common and proposed methods of predicting roughness in the ripple and sheet flow regimes, and presents results of evaluating the theory and methods with field data. The new theory combined with either roughness method successfully predicts current shear velocities in wave-current field flows over beds in the lower flat-bed, ripple, and sheet flow regimes, with the proposed method yielding the smaller bias. Remaining questions concerning the appropriate near-bottom orbital velocity required to describe field conditions must be resolved when additional field data becomes available.