ABSTRACT. Local reservoir ages are often estimated from the difference between the radiocarbon ages of aquatic material and associated terrestrial samples for which no reservoir effect is expected. Frequently, the selected aquatic material consists of bivalve shells that are typically well preserved in the archaeological record. For instance, large shell middens attest to the importance of mussel consumption at both coastal and inland sites. However, different physiological mechanisms associated with tissue and shell growth may result in differences in reservoir effects between the surviving component (shell) and the component relevant to dietary reservoir effects in consumers (tissue). The current study examines bivalve tissue-shell age differences both from freshwater and marine contexts close to archaeological sites where human consumption of mollusks has been attested. Results exhibited significant ,4 C age differences between bivalve tissue and shell in a freshwater context. In a marine context, no significant bivalve tissue-shell age differences were observed. The results also showed that riverine and lacustrine shells show large and variable freshwater reservoir effects. The results have important implications for establishing local reservoir effects especially in a freshwater environment. For good a priori knowledge of expected 14 C differences in organic and inorganic water, carbon is thus necessary. Furthermore, the high variability in freshwater shell 14 C ages implies the need for representative sampling from the archaeological record.