Abstract The Dyfed Supergroup of SW Wales ranges from Terreneuvian to Furongian in age and represents a prolonged and nearly continuous phase of siliciclastic sedimentation on the southern margin of the Cambrian Welsh Basin. The lower 564 m of the Supergroup are included in an extended Caerfai Group (six formations), embracing strata previously assigned to the Solva Group. Predominantly arenaceous, the Caerfai Group has important units of conglomerate and pebbly sandstone at its base, middle and top. The Caerfai Group is overlain by 687 m of mainly argillaceous sedimentary deposits of the Porth-y-rhaw Group (five formations), a newly erected unit that includes strata previously assigned to the ‘Menevian Group’ and ‘Lingula Flags’. This two-fold division of the Dyfed Supergroup is comparable with the distinction of the laterally equivalent Harlech Grits and Mawddach groups exposed around the Harlech Dome in North Wales. High resolution sequence-stratigraphical techniques, constrained by biostratigraphical data wherever possible, are applied to the Dyfed Supergroup across southern Britain: the revised lithostratigraphy is thereby integrated with a slightly modified version of the Avalonian chronostratigraphy initially developed for the western Avalonian successions of maritime Canada. Sequences 1 and 2 are not recognized in the Dyfed supergroup of SW Wales, reflecting the inner platform setting of the area. Those sequences are represented in North Wales however, which was situated on the outer platform. The relative ease with which the Cambrian successions of southern Britain can be incorporated into the general Avalonian framework reflects the shared epeirogenic history, sediment sources and accumulation history of a microcontinent unified by early Cambrian times. The gross lithological similarities that exist between Cambrian successions across Avalonia, and between SW Wales and North Wales in particular, are better understood when basin chronostratigraphy is placed within a sequence stratigraphic framework defined by systems tracts. This is particularly evident during Unnamed Cambrian Series 3 (C3) and Furongian times, when clear systems links are demonstrated between paralic depositional environments in SW Wales and deeper basin turbidites in North Wales. Sequences 3–7 (Terreneuvian–C3) in the Welsh Basin and its hinterland were dominated by siliciclastic deposition on an epeirogenically active platform. An extensional rift-like tectonic regime is proposed, where elongate basins dominated by mudstone deposition have time equivalents formed on areas of apparent uplift that were probably tilt-block highs. Local accumulation history, the development of regional unconformities, the extent of marine onlap, and the secular succession of lithofacies were controlled principally by phases of fault-accommodated subsidence along the NE-trending lineaments that bounded these basins and their intervening horsts or platforms. In southern Britain, movements first along the Menai Straits Fault System then the Welsh Borderlands Fault System, both of which are terrane boundaries, were responsible for the dominantly coarse-grained arenaceous sedimentation of pre-Drumian times. In the Welsh Basin, this is highlighted by a change in sediment source from the Monian Superterrane to the Wrekin Terrane. A rapid switch from marginal to inner platform source areas accompanied a major phase of tectonic reorganization of the Avalonian Superterrane during the development of the boundary between sequences 3 and 4. Although the role of secular variations in eustatic sea level as a control on sequence development and architecture has been dismissed previously, prominent lowstands such as those recorded during the basal Ptychagnostus gibbus Biozone, ‘ Solenopleura ’ brachymetopa Biozone (mid- Paradoxides forchhammeri Biosuperzone), and the upper part of the forchhammeri Biosuperzone also influenced the depositional sequence stratigraphy. The broad subdivision of Cambrian Avalonian stratigraphy into tectonically active and passive phases of sedimentation allows two megasequences to be distinguished. Megasequence 1 (sequences 1–7) represents the transform termination of Avalonian subduction following oblique convergence, and the accretion of island arcs onto the northern periphery of Gondwana. Under transpressional regimes, late Precambrian arc-related basins were inverted and recycled into pull-apart successor basins. Initially, the margins of these were dominated by alluvial fans and coarse-grained fan-deltas built by flood-generated sediment dispersal processes. In the later transform stage of Cadomian–Avalonian orogenesis, there was a switch to sediment supplied from highlands to the east. Sequence 8, represented in Pembrokeshire by the Aber Llong and Ogof Velvet formations, lies at the base of Megasequence 2 (late C3–early Ordovician). Excellent sections occur through these formations and their equivalents in North Wales, and their interpretation significantly improves understanding of younger Cambrian sedimentary basins in Avalonia. These successions reflect passive margin sedimentation and the culmination of the Avalonian–Cadomian orogenic cycle. Deposition occurred in part of a subaqueous delta platform at the mouth of a huge river system that drained a substantial part of West Gondwana, with Avalonia acting as a sink for vast quantities of fine-grained sediment. Secular variations in eustatic sea level and/or sediment supply, rather than active tectonism, were the main mechanisms controlling sequence architecture and depositional environments at this time.