Given the current shale oil boom in U.S., the primary objective of this dissertation is to clarify the factors controlling retention and migration of oil in selected shale ”plays”. With reference to three TypeII marine shales namely the Barnett, Posidonia and Niobrara Shales, a suite of 694 samples ranging in age from Mississippian to Campanian, varying in maturity from immature to overmature, and belonging to siliceous, argillaceous and calcareous lithofacies were comprehensive investigated. Beginning with the study of oil retention, a more realistic assessment of total retained oil was achieved by applying the comparatively pyrolysis: Total oil = S1whole rock + S2whole rock - S2extracted rock. Notably, as excellent correlations always exist between the calculated total oil with the S1 values of unextracted rocks at given levels of maturity, the S1 was used in the ensuing discussions as a proxy to reflect total oil concentration. Oil retained in shales is either in a sorbed state largely on kerogens or in a free form in pores and fractures. In organic-rich shales, the retention of oil is defined mainly by organic matter properties, i.e. organic richness, kerogen type and thermal maturity. The sorption behavior of kerogen is believed analogous to that of organic polymers, which are capable of absorbing significant quantities of oils by swelling. Following this hypothesis: 1). The richer in organic matter a rock is, the more is the oil that is retained. 2). Labile kerogens, rather than inert carbon, constitute the active swelling components. 3). For TypeII marine source rocks, S1/TOC first increases and subsequently decreases once the maximum retention capacity (90 mg HC/g TOC) is exceeded at Tmax about 445 °C, which is equivalent to ~ 0.85% Ro. But interestingly, the shale layers enriched in free oil or bitumen are not necessarily associated with the layers richest in organic matter, and instead with juxtaposed porous biogenic matrices. In the siliceous interval of the Barnett Shale oils are thus stored in the axial chamber of sponge spicules. In the Posidonia Shale, bitumen was observed in pores of associated coccolith microfossils. In the chalky reservoirs of the Niobrara Formation it is carbonate richness that primarily controls the amount of retained oil (S1). Oils are mainly stored in pores associated with the skeletal remains of coccolith and foraminifera. These porous fossiliferous layers in shale may constitute sweet spots (reservoirs) due to their enhanced hydrocarbon potential and mobility. In contrast to clastic reservoirs, the oil-in-place of shale reservoirs may be either indigenously generated or migrated from juxtaposed organic-rich layers. According to the studies of the Barnett, Posidonia and Niobrara Shales, the presence of an “oil crossover” and diminished Tmax are likely to be characteristic of those fossiliferous shale reservoirs. To settle this issue in the Barnett Shale of the Mesquite#1 well, a mass-balance model was used to compute the hydrocarbons generated. By comparing the amount with that retained, it is clear that more hydrocarbons (C13+n-alkanes, such as the n-C17) are in-place than could have been generated. Therefore, additional hydrocarbons must have migrated and accumulated in the siliceous interval, i.e. reservoirs. In the Mesquite#1 well, the Barnett Shale was shown to possess a rather homogeneous kerogen facies, depositional environment and maturity signature, whereas the composition of bitumens varied throughout the shale sequence. The short distance migration of petroleum into the siliceous reservoir interval appears to fractionate the generated oil into a higher quality oil by preferential retention in the order polar compounds >aromatic >saturated hydrocarbons within the underlying organic matter- and clay-rich source rock intervals. Besides that, a preferential expulsion of smaller molecular components over larger ones, i.e. molecular fractionation, has been illustrated. Phase separation is a possible scenario leading to molecular fractionation. As phase separation selectively “transfers” lighter hydrocarbons into the vapor phase, the oil retained in migration pathways or reservoirs is otherwise enriched in retrograde fluids with n-alkanes skewed towards long chain alkanes. Regularities in source facies and maturity were confirmed with biomarkers, which are not fractionated during the primary migration of petroleum within the Barnett Shale of Mesquite#1 well. For given components, if migrational fractionations had occurred, they might have been overprinted by in situ reactions as well. Organic pore development is believed to be largely due to the thermal cracking of kerogen and/or bitumen, though some primary organic pores have been observed within immature organic matter as well. Oil retention and organic porosity evolution are strongly related to changes in kerogen density brought about by swelling and shrinkage as a function of thermal maturation. For TypeII marine shales, the secondary organic pores are formed consequently after the maximum kerogen swelling ability is exceeded at Tmax around 445 °C (~0.85% Ro). Shrinkage of kerogen itself leads to the formation of secondary organic pores, and thence associated porosity increase in the gas-mature Posidonia Shale. Given the remarkable heterogeneity in the shale fabric, the newly formed organic pores may be closed after or synchronously by compaction.