Time-averaging has evolved from an unrecognized variable in paleoecological analyses to a key concept in understanding the dynamics of sedimentary systems and the formation of fossil deposits. Here we used radiocarbon-calibrated amino acid racemization ages from 173 Fulvia tenuicostata shells collected from Sydney Harbour (NSW, Australia) to quantify time-averaging in surficial and excavated death assemblages. A novel approach to estimating the total age-estimate variability of a collection of specimens is presented that integrates the age-estimation error associated with the AAR calibration model and the effect of time-averaging variability on the age distribution. Fulvia collected from a single 1.6-m deep excavation were used to quantify changes in time-averaging with burial depth. Fulvia collected from surficial death assemblages at six sites were used to quantify spatial variation in time-averaging. The median shell age increased from ∼ 150 yr to ∼ 4230 yr and time-averaging from ∼ 40 yr to ∼ 960 yr with increased burial depth. While four sites contained surface shell assemblages with median ages of ∼ 150 yr and time-averaging of ∼ 40 yr, two sites had death assemblages with older median ages and time-averaging > 1900 yr. A decline in the abundance of Fulvia in the post-colonial period is likely responsible for the rarity of very young shells in the surface death assemblages, while local factors such as dredging and other human activities are likely responsible for the total age-estimate variability in the two highly time-averaged surface collections. These analyses provide a geochronological context for assessing ecological changes in the harbor's benthic communities since colonization and offer an important perspective on the formation of Holocene fossil deposits.