Understanding and predicting the size, shape and internal architecture of sandbodies is of fundamental importance in hydrocarbon exploration. High quality hydrocarbon reservoirs are often formed in deltaic environments where there is a complex interplay between changes in relative sea level, sediment input and climate. When combined with the intrinsic sedimentary variability of deltas, this makes prediction of the internal facies distributions and architectures of deltaic sandbodies a challenging task. The aim of this thesis is to conduct a detailed qualitative and quantitative field study of fluvial and deltaic sandbody architecture and facies distribution, and to assess the usefulness of the data thus obtained in predicting the architecture and sedimentary characteristics of reservoir sandstones in the subsurface. The El Wastani Formation of the Nile Delta, Egypt, has previously been identified as an interval of reservoir quality sandstones within the Plio-Pleistocene deltaic succession. Limited core data, and poor seismic imaging due to gas seepage effects, hindered past attempts to assess the internal architecture and facies of the sandbodies. Therefore it was considered appropriate to use an outcrop analogue to aid understanding of the El Wastani Formation reservoir characteristics. From a review of literature, the Upper Carboniferous fluvial and deltaic sandstones of the Central Clare Group, County Clare, western Ireland, were found to be suitable analogues for the El Wastani Formation sandstones. Controls on the two sedimentary systems were similar; both were fluvial-dominated and wave-influenced, and both show evidence for fluctuating relative sea-level. Comparisons of facies observed in outcrop (Co. Clare) and interpreted from image logs (Nile Delta) show similar facies and sedimentary successions in the two systems, improving confidence in the choice of analogue. Fieldwork carried out on the Upper Carboniferous (Namurian) coastal outcrops of Co. Clare produced detailed measurements of facies distributions and bed geometries, which, together with sedimentary logs, palaeocurrent studies and outcrop-scale photomontages, enabled interpretation and quantification of channel dimensions, internal architectures and stacking patterns. Based on these data, the Tullig Sandstone, a major sandbody within the Central Clare Group, is interpreted to be a low-sinuosity, braided fluvial system that flowed to the north-northeast. This sandbody shows decreasing amounts of erosion and conglomeratic facies in both downstream and vertical directions, interpreted to reflect the combined effects of delta subsidence and sea-level rise over time, influencing the downstream reaches of the system first. The mean sand to non-sand ratio for the Tullig Sandstone is 97% by area, and connectivity of sandstone facies within this sandbody is 93%. In contrast, mouthbar sandbodies that were studied have a mean sand to non-sand ratio of 90%, and greatly reduced sandstone connectivity, at 65%. The data that characterise the field outcrops can be taken as indicative of the probable characteristics of the El Wastani sandbodies. The data generated from the quantitative field studies were used to construct computer models of the outcrops, in order to see how well the modelling software was able to reproduce the outcrop architectures and facies distributions, and also to test the sensitivity of the models to different scales of data. One large-scale model was built to include all the Tullig Sandstone outcrops along the coastline, with a vertical resolution (cell height) of 1m. A second smaller model was constructed to cover just the Trusklieve outcrop, and was built using a vertical cell height of O.1m. Each model was designed to fit the sedimentary log data, and was conditioned to reflect the facies percentages and channel dimensions measured and calculated respectively from the outcrops. The results showed that although the larger modelling grid, with lower vertical and horizontal data resolution, showed significant differences in finegrained facies distribution from the outcrops, it was reasonably successful at reproducing the channel shapes and stacking patterns seen in outcrop. In addition, the high sand to non-sand ratio meant that sandstone connectivity was not reduced compared with either the outcrops or the small, high-resolution model. The small model was better at reproducing the geometries of beds of fine-grained facies, but lacked the ability to accurately simulate the channel architectures and stacking patterns.