Abstract The Cretaceous Mesaverde Group of the Bighorn Basin, northwestern Wyoming, is comprised of two major clastic wedges that record the progradation and retrogradation of deltaic depositional systems within the Cretaceous Western Interior Seaway. Within the Campanian Virgelle and Judith River Formations 16 sand-rich clastic tongues, deposited in mixed wave/storm-dominated shallow marine shoreface depositional environments, have been studied and traced into equivalent updip non-marine and downdip offshore facies. Each tongue is typically a massive shoreface sandbody that pinches-out and correlates basinward (east) with fine-grained offshore heterolithic progradational parasequences. Regional correlation reveals the sandstone tongues to be sharp-based. Their lower bounding surfaces are characterized by: (1) a marked basinwards shift in facies, (2) an abrupt increase in sand: shale ratio, (3) missing/eroded facies below, (4) a change in parasequence stacking patterns, (5) local development of Glossifungites firmground ichnofabrics, (6) deposition of precursor gutter cast facies, (7) widespread soft-sediment deformation and growth faulting, (8) changes in palaeocurrent orientation, (9) regional truncation of older parasequences and systems tracts, (10) regional depositional-dip correlation of 20–40 km. The basal sharp-based surfaces of the shoreface sandstones are interpreted to be regressive surfaces of marine erosion (RSME) formed during falls in relative sea-level, with the shallow marine successions deposited during forced regression of the shoreline. Internal hetrogeneities and erosion surfaces within the massive shoreface sandstones are interpreted to record stepwise progradation during relative sea-level falls. These relatively steep seaward-dipping erosion surfaces reflect the overall trend of shoreface deposition and amalgamation during periods of decreasing accommodation space. Downdip, the internal erosion surfaces amalgamate with the basal RSME and provide evidence that this basal surface is composite in nature and as such diachronous in its development. Within the studied interval, four examples of forced regressive deposits are confidently correlated updip to correlative incised valley fills. In each case, the basal erosion surface to the incised valleys truncates strandplain deposits, and ties laterally to a subaerial exposure surface (interfluve) developed across the top of the strandplain. These surfaces, formed in response to a fall in relative sea-level, are interpreted as sequence boundaries. Traced basinward their expression is commonly lost as the upper strandplain and capping interfluve are eroded by transgressive ravinement at the base of tidal inlets. However, the interfluve is thought to correlate downdip to the final seaward dipping erosion surface that separates the massive amalgamated shoreface sandstones of the falling stage system tract, deposited during overall relative sea-level fall, from the more heterolithic parasequences of the low-stand systems tract, deposited under conditions of stillstand to relative sea-level rise. Within sediments deposited during relative sea-level fall the critical transition from the subaerial to submarine expression of the sequence boundary is recognized as the main factor in the ongoing controversy regarding the identification of a finite chronostratigraphic sequence-bounding surface. Drawbacks exist in making a simple choice between the subaerial exposure surface or the RSME as the sequence boundary because they are normally diachronous and at the same time form contemporaneously. In updip areas two separate important stratigraphic surfaces may be distinguished; the RSME and an overlying subaerial exposure surface. In these areas the subaerial exposure surface must be regarded as the main sequence bounding surface. Basinward exists a critical transition zone where (i) storm-related erosion surfaces and shoreface amalgamation during deposition of the strandplain inhibit correlation and (ii) transgressive ravinement may erode part of the subaerial expression of the sequence boundary. In this area choice of surfaces proves difficult. In constrast, basinward of the last sharp-based shoreface, the RSME would be the principal (and most obvious) stratal surface, and must be interpreted as a sequence boundary. By considering the evolution of the RSME to occur at the same time as the fluvial erosion/subaerial exposure surfaces, the massive sharp-based shorefaces and their distal equivalents can be observed to be the response of an linked and dynamic system to relative sea-level fall.