A major challenge in sequence stratigraphy is objectively identifying stratigraphic surfaces and sequences across multiple scales of observation. Identification is commonly dependent on the resolution of the data used (i.e., seismic vs. well data), its dimension (i.e., 1D vs. 2D vs. 3D) or the criteria chosen to select sequence boundaries. Through shelf-edge trajectory analysis, the clinothem (i.e., highest order seismic sequence identified on seismic data) can constitute the elementary building block of an observation-based and data-driven quantitative workflow to develop sequence stratigraphic frameworks across different orders and ranks of hierarchy. Here, we use high-quality 3D seismic data to interpret a Late Tithonian–Early Cretaceous shelf margin, the Lower Barrow Group (LBG), developed in the Northern Carnarvon Basin on the North West Shelf of Australia. Based on full-volume seismic interpretation techniques that integrate the 3D variability of the data when identifying seismic unconformities, a high-resolution seismic stratigraphic framework was built (73 interpreted clinothems with an average time duration of ~63,000 yrs). The computation of high-frequency shelf-edge trajectory angle (Tse) curves on selected seismic cross-sections is used to objectively pick sequence stratigraphic surfaces based on the accommodation succession method, thereby highlighting small changes in trajectory and proposing a method reproducible by interpreters based on the same quantitative data. Within the D. lobispinosum interval (142.3–140.9 Ma), the definition of stratigraphic sequences and composite stratigraphic sequences through this workflow is used to discriminate the controls at high and low temporal frequency on the vertical and lateral variability (which is here quantified) of this shelf-slope-basin system. The results show that the high-frequency interplay between short-term glacio-eustasy (i.e., Milankovitch eccentricity cycles of ~100,000 yrs) and sediment supply (locus of fluvial input along the margin) impacted the three-dimensional stratigraphic architecture of the LBG. In contrast, tectonic subsidence had a significant impact on the stratigraphic architecture of the LBG within the main depocentre at lower temporal frequency by overprinting the eustatic signal and accelerating/decelerating the rates of accommodation creation. However, identification of long-term glacio-eustatic Milankovitch cycles (~400,000 yrs) outside the main depocentre, where the rates of accommodation creation due to rift-related subsidence are moderate, also suggests low-frequency eustatic control. Therefore, the vertical and lateral variability of the LBG results from variations in sediment supply and subsidence regime under local (i.e., process regime, currents), regional (i.e., tectonics) and global (i.e., eustasy, climate) forcing parameters interplaying across timescales. In contrast to standard sequence stratigraphic workflows that are based on model-dependent choices to select sequence boundaries, quantitative 3D seismic stratigraphy constitutes an improved method to interpret 3D seismic data in shelf-margin depositional systems within a sequence stratigraphic framework, which provides an observation-based and model-independent tool allowing the definition of stratigraphic sequences with results that are reproducible across multiple stratigraphers. This work highlights the need for developing new sequence stratigraphic tools and methods that integrate the 4D variability of depositional systems and moves beyond the two-dimensionality inherent to current sequence stratigraphic methods. Quantitative 3D seismic stratigraphy represents a first step towards the creation of 3D sequence stratigraphic workflows that could improve the prediction of stratigraphic patterns and facies relationships (source, reservoir, seal distribution) across shelf margins.