Oxygen-isotope measurements of fossil carbonates remain the most common method for paleoclimatic temperature reconstructions. A well-known limitation of this approach is the influence of the oxygen isotope composition of water in which mineralization occurs, which may vary significantly through space and time, and is often difficult to constrain precisely. Carbonate clumped-isotope thermometry is an alternative approach applicable to many carbonates. It is based on measurements of Δ47 (a tracer of small statistical anomalies in the abundance of rare, doubly-substituted carbonate isotopologues), and requires no independent information on the oxygen-isotope composition of parent waters. Here, we report new calibration observations of clumped isotopes in four species of calcitic marine bivalves (A. colbecki, N. cochlear, S. cucullata, M. gigas) from various ecosystems including coastal and deep-sea environments, with calcification temperatures ranging from −2 °C to 27 °C and very different amplitudes of seasonal temperature variability. At two localities with large seasonal temperature variability, calcification time intervals were constrained using a sclerochronological approach to test whether seasonal gradients of temperature can be accurately quantified based on Δ47 measurements.Our results indicate that the mature bivalves we analyzed have clumped-isotope compositions entirely consistent with earlier calibration studies processed in the I-CDES reference frame and based on biogenic/abiotic/synthetic materials. By contrast, juvenile M. gigas oysters yield substantially lower Δ47 values than expected based on their calcification environments, suggesting that their early growth phase is associated with yet poorly understood isotopic biases affecting both δ18Ο and Δ47 values. The link between seawater temperatures and bivalve Δ47 values is thus potentially applicable to seasonal reconstructions, but only if shell sections formed in cold seasons are precisely identified and precisely sampled, and taking into account that winter calcification is likely to be biased due to reduced growth rate. Moreover, the excellent agreement between our observations and the existing I-CDES calibrations further demonstrates the efficacy of the I-CDES standardization approach, and adds to the evidence that many different types of carbonates conform to statistically indistinguishable relationships between Δ47 and crystallization temperature.