Abstract The oxygen and strontium isotopic ratios of marine carbonates are widely employed to derive paleotemperature and age, respectively. While open ocean samples are the standard source for these measures, shelf settings can also provide detailed records. However, the implementation of shelf data for these purposes has been hindered by the perception that local environmental conditions (such as freshwater influx) confound regional and global signals. Here, we directly evaluate this concern by assessing the spatial variability in the isotopic composition of biogenic carbonate precipitated along the depositional strike of an inner shelf environment. Data are derived from mollusc shells collected from a stratigraphically restricted, transgressive unit in the early Eocene (~55Ma) of the US Gulf Coastal Plain (GCP). Lithofacies and fossil assemblages suggest that an estuary was present in the westernmost study region, whereas more typically marine facies predominated elsewhere. Isotopic composition varies predictably with paleoenvironmental setting; shells from the estuary have δ18O values and 87Sr/86Sr ratios that reflect high freshwater input. Specimens in the eastern GCP, however, from within and between outcrops separated by up to 400km, all yield similar results. Serial microsampling reveals congruence not only in mean δ18O values, but also in winter and summer extremes. Estimated paleotemperatures (mean annual temperature of 26°C) agree well with proxy and model evidence for the early Eocene and the GCP, and 87Sr/86Sr ratios are consistent with the established early Eocene seawater ratio. The fidelity of isotopic values relative to environment indicates that shallow shelf records have much to contribute to studies of ancient climate and ocean chemistry.