A transect of four sites (U1325, U1326, U1327 and U1329) across the northern Cascadia margin was established during Integrated Ocean Drilling Program Expedition 311 to study the occurrence and formation of gas hydrate in accretionary complexes. In addition to the transect sites, a fifth site (U1328) was established at a cold vent with active fluid flow. The four transect sites represent different typical geological environments of gas hydrate occurrence across the northern Cascadia margin from the earliest occurrence on the westernmost first accreted ridge (Site U1326) to the eastward limit of the gas hydrate occurrence in shallower water (Site U1329). Expedition 311 complements previous gas hydrate studies along the Cascadia accretionary complex, especially ODP Leg 146 and Leg 204 by extending the aperture of the transect sampled and introducing new tools to systematically quantify the gas hydrate content of the sediments. Among the most significant findings of the expedition was the occurrence of up to 20 m thick sand-rich turbidite intervals with gas hydrate concentrations locally exceeding 50% of the pore space at Sites U1326 and U1327. Moreover, these anomalous gas hydrate intervals occur at unexpectedly shallow depths of 50-120 metres below seafloor, which is the opposite of what was expected from previous models of gas hydrate formation in accretionary complexes, where gas hydrate was predicted to be more concentrated near the base of the gas hydrate stability zone just above the bottom-simulating reflector. Gas hydrate appears to be mainly concentrated in turbidite sand layers. During Expedition 311, the visual correlation of gas hydrate with sand layers was clearly and repeatedly documented, strongly supporting the importance of grain size in controlling gas hydrate occurrence. The results from the transect sites provide evidence for a structurally complex, lithology-controlled gas hydrate environment on the northern Cascadia margin. Local shallow occurrences of high gas hydrate concentrations contradict the previous model of gas hydrate formation at an accretionary prism. However, long-lived fluid flow (part of the old model) is still required to explain the shallow high gas hydrate concentrations, although it is most likely not pervasive throughout the entire accretionary prism, but rather localized and focused by the tectonic processes. Differences in the fluid flow regime across all of the transect drill sites indicate site-specific and probably disconnected (compartmented) deeper fluid sources in the various parts of the accretionary prism. The data and future analyses will yield a better understanding of the geologic controls, evolution and ultimate fate of gas hydrate in an accretionary prism as an important contribution to the role of gas hydrate methane gas in slope stability and possibly in climate change. ?? The Geological Society of London 2009.