Abstract Coastal fluvial and deltaic deposits are locally incomplete recorders of elapsed time, but where detailed three-dimensional stratigraphic geometries can be constructed from well sections or high-resolution seismic data, and dated using high-precision chronostratigraphic methods, such as 14C dating, the construction of three-dimensional time-space (Wheeler) diagrams reveals important insights regarding the relationships between basin architecture, accommodation rates, sedimentation rates, stratigraphic preservation, and the nature of autogenic and allogenic processes. Analysis of the time-stratigraphic history of point-sourced delta systems, using strike and dip cross sections, is used to evaluate the completeness of the stratigraphic record to resolve how basins fill, how vertical stratigraphic successions record time, and which surfaces record the longest hiatuses. The paper includes analyses and comparison of Quaternary and Cretaceous river and delta systems. Supply-dominated systems fed by rivers that are free to avulse, like the Holocene highstand Mississippi Deltas in the Gulf of Mexico, show a strong autogenic record. In the Mississippi deltas, a quasi-continuous Holocene stratigraphic record can be reconstructed but this record is distributed among a series of laterally offset delta complexes and component delta lobes. Although 1D vertical sections may sample less than 25% of the total time, and in many cases less than a few percent, a hiatus in one section typically correlates to deposits elsewhere. Deltas fed by entrenched rivers, and especially the lowstand shelf edge systems in the Gulf of Mexico, as well as the Mahakam in Kalimantan Indonesia, form a series of laterally-offset lobes at the shelf edge that look superficially similar in their pattern and areal scale to the autogenic lobes of the Mississippi delta, but the time stratigraphic relationships are very different. Rather than a quasi-complete time-stratigraphic record, each shelf edge lobe is interpreted as an allogenic lowstand delta, deposited over tens of thousands, versus thousands of years, and are bounded by a prolonged non-depositional hiatus, time equivalent to highstand deposition confined to the inner shelf. Analysis of strike cross sections within an allogenic lobe of the Lagniappe lowstand delta reveals an autogenic signal, similar, but at a smaller scale than the Mississippi. In 1D, the progradational part of an upward coarsening deltaic parasequence records the time it takes the delta to prograde past a given point, which is estimated to be a few hundred years, representing only a few percent of the time associated with progradation and transgression of a parasequence. Nevertheless, this part of the succession records the highest rates of deposition (typically 10–50 m/ka). Much of the time associated with a parasequence is tied up at the top surfaces that record periods of bypass, as the lobe progrades farther seaward, as well as transgressive surfaces of erosion and deepening. The trangressive units deposit at much slower rates (1 m/ka) recorded by highly bioturbated condensed sections. Analysis of fluvial and deltaic floodplains and channel belts also show a highly incomplete record (as little as 5% of elapsed time), reflecting the observation that rivers only occupy a small part of any given floodplain at any given time, especially when rivers are entrenched and floodplains may be exposed to form paleosols. Entrenched river deposits, such as valley fills, may record only a few percent of the total time, indicating that most of the valley records bypass and erosion. Aggrading floodplains, which may be more common during highstands of sea level, may contain more complete records, including peats and coals, with up to 50% of time recorded by sediment. A tentative reconstruction of the development of the fluvial Canterbury Plain of South Island, New Zealand, is offered as an example of the highly discontinuous nature of alluvial sedimentation.