The South Bay Salt Pond Restoration Project is overseeing the restoration of about 6000 ha of former commercial salt-evaporation ponds to tidal marsh and managed wetlands in the southern reach of San Francisco Bay (SFB). As a result of regional groundwater overdrafts prior to the 1970s, parts of the project area have subsided below sea-level and will require between 29 and 45 million m³ of sediment to raise the surface of the subsided areas to elevations appropriate for tidal marsh colonization and development. Therefore, a sufficient sediment supply to the far south SFB subembayment is a critical variable for achieving restoration goals. Although both major tributaries to far south SFB have been seasonally gaged for sediment since 2004, the sediment flux at the Dumbarton Narrows, the bayward boundary of far south SFB, has not been quantified until recently. Using daily suspended-sediment flux data from the gages on Guadalupe River and Coyote Creek, combined with continuous suspended-sediment flux data at Dumbarton Narrows, we computed a sediment budget for far south SFB during Water Years 2009–2011. A Monte Carlo approach was used to quantify the uncertainty of the flux estimates. The sediment flux past Dumbarton Narrows from the north dominates the input to the subembayment. However, environmental conditions in the spring can dramatically influence the direction of springtime flux, which appears to be a dominant influence on the net annual flux. It is estimated that up to several millennia may be required for natural tributary sediments to fill the accommodation space of the subsided former salt ponds, whereas supply from the rest of the bay could fill the space in several centuries. Uncertainty in the measurement of sediment flux is large, in part because small suspended-sediment concentration differences between flood and ebb tides can lead to large differences in total mass exchange. Using Monte Carlo simulations to estimate the random error associated with this uncertainty provides a more statistically rigorous method of quantifying this uncertainty than the more typical “sum of errors” approach. The results of this study reinforce the need for measurement of estuarine sediment fluxes over multiple years (multiple hydrologic conditions) to adequately detail the variability in flux. Additionally, the timing of breaching events for the restoration project could be tied to annual hydrologic conditions to capitalize on increased regional sediment supply.