Understanding the origins of shelf-edge deltas is important because they area key component in the basin-scale sediment routing system. Previous analyses have shown that shelf-edge deltas may be either lowstand or highstand features, depending on the magnitude of fluvial sediment supply, rate of accommodation creation, and the operation of delta autoretreat. In this paper, a stratigraphic forward model is used to generate progradational sediment wedges of the type found oil many basin or shelf margins. Numerical experiments are conducted to further investigate how delta autoretreat influences shelf-edge delta formation, and to investigate relationships between sediment supply, accommodation creation, and formation of shelf-edge deltas generally. Results from the stratigraphic forward modeling are limited by the assumptions inherent in the model, but they suggest that autoretreat is more likely in systems with relatively high rates of marine sediment transport. Autoretreat appears to be least effective at low rates of sea-level rise, high rates of sediment supply, and low rates of marine sediment transport. Model results also suggest that sediment-wedge topset width (shelf width plus width of coastal plain) is controlled by the initial bathymetry into which delta progradation occurs, and the balance between sediment Supply and long-term rate of accommodation creation, with higher-frequency relative sea-level oscillations playing only a minor role once the topset width has been determined by initial progradation. Based on this finding, formation of highstand shelf-edge deltas may be a process of self-regUlated equilibrium regression; new phases of progradation, either forced or unforced, can reach the shelf edge by default because the shelf-edge position is a consequence of previous phases of progradation. Thus well-supplied shelf-edge deltas fed by medium to large river systems with consistent sediment supply may commonly form somewhat independently of forcing by relative sea-level oscillations, implying that throughout much of the geological record, timing of delivery of sand to deep-marine systems is unlikely to depend on simple forcing by short-term relative sea-level oscillations.