A large integrated data set of cores, outcrop data, and seismic transects from the mud-buried Vars-Winchester esker in the Champlain Sea basin, Canada, was studied to gain insight into how muddy glaciated basins fi ll with sediment, and how esker sedimentary systems contribute to this process.Three stratigraphic units-a till sheet over carbonate bedrock, the Vars-Winchester esker , and overlying Champlain Sea mudare identifi ed in the data set. The till is massive, mud rich, carbonate rich, and drumlinized. The esker is also carbonate rich, and rests erosively on till or bedrock. It consists of two elements, a narrow gravelly central ridge and a broad sandy carapace. Three units comprise the overlying mud package: gray carbonate-rich rhythmites, massive bioturbated mud, and carbonate-poor, red-andgray rhythmites. A sequence stratigraphic model is proposed to explain these observations. Emphasis is placed on gradual ice-front translation superimposed by rapid meltwater events. The esker is interpreted to have been derived from the underlying till by water that fl owed through a subglacial conduit (R-channel), within which the narrow gravelly central ridge was deposited. Most mud and fi ner sand bypassed the conduit and was deposited proglacially on the fl oor of the Champlain Sea, fi rst as sandy outwash and, farther basinward, as muddy carbonate-rich rhythmites. Gradual ice-front retreat superposed distal facies over proximal facies, generating the upward-fi ning succession that starts with the esker gravel and ends with muddy rhythmites. Most esker sediment appears to have been deposited during rapid, jökulhlaup-like fl oods that punctuated grad-ual retreat. Discharges are estimated to have been high, possibly on the order of several hundred to, perhaps more commonly, several thousand cubic meters per second. The chaotic and random-looking appearance of the resultant sedimentological signatures in the esker sensu stricto is sharply contrasted with the regularity of the muddy rhythmites. If the rhythmites are indeed correlative to the esker, which seems reasonable given their geochemistry and the fact that their volume scales to the volume of mud in the till, the fl ood events that deposited the esker must have been seasonally mediated, and the basin water must have attenuated the fl ood signal, resulting in a rhythmic "on-off" signature in more distal portions of the system. The regularity of the rhythmites does not betray the chaotic nature of the esker sensu stricto, and vice versa. Studying either one in isolation would lead to a very different "end-member" impression of how eskers form and how esker sedimentary systems operate during the infi lling of glaciated basins.
