Primary petroleum migration (expulsion from source rocks) remains the least understood parameter controlling the genesis of oil deposits. In spite of this lack of understanding, many petroleum geochemists (including this Author) have previously considered expulsion from organic-rich, mature source rocks to be very efficient. This viewpoint results from Rock-Eval analyses of organic-rich source rocks, analyses which demonstrate a loss of hydrocarbon (HC) generation capacity, by significant reduction in Rock-Eval hydrogen indices, as such rocks are progressively buried in sedimentary basins. However, this progressive loss of HC generation capacity is not matched by numerically-equivalent increases either in Soxhlet-extractable HCs or the Rock-Eval S1 pyrolysis peak. Thus, we conclude that almost all generated HCs have migrated from the source rocks. Notwithstanding, the petroleum geochemistry of the Williston Basin (North America) and other considerations strongly suggest that this logic may be flawed. Instead, it appears that most generated HCs may not migrate far from their generation site, but instead are lost before source-rock samples arrive at the laboratory for analysis. This loss occurs from a HC-gas volume expansion, which results from the large pressure decreases in rock chips or cores during the trip up the wellbore in the course of drilling operations. The large volume expansions of these HC gases, which are cogenerated and coexist with oil in the source rocks, literally blow most generated oil in source rocks into the drilling mud during the trip uphole. If most generated HCs do in fact remain in or near their source rocks, then it can be hypothesised that source rocks must be physically disrupted before meaningful expulsion can occur. Faulting, with accompanying significant fracturing, would appear to be the optimum naturally-occurring process for physical disruption of source rocks. If these hypotheses are valid, intensity of faulting in deeply-buried HC ＇＇kitchens＇＇ containing mature source rocks should strongly correlate with increasing basin richness, as defined by recoverable oil divided by basin-sediment area or volume. This possible relationship is examined in this Paper; and there is a strong correlation of increasing basin richness with increasing structural intensity over and adjacent to basin depocentres. This correlation thus supports the hypothesis that physical disruption of mature source rocks is a necessary, and previously unappreciated, controlling parameter for oil expulsion. If a relationship between physical disruption of source rocks and oil expulsion indeed exists, then significant implications would follow for: (1) basin resource assessment; (2) conventional oil exploration in frontier basins; and (3) the probable existence of very large, previously unappreciated, oil-resource bases in fractured, self-sourced shales.