Uptake of increasing anthropogenic CO2 emissions by ocean surface waters is causing an increase of seawater PCO2 accompanied by a decrease of seawater pH and carbonate ion concentrations. This process, termed ocean acidification (OA), is predicted to negatively affect many marine organisms with likely consequences for marine ecosystems and the services they provide. Calcifying mussels and oysters, and particularly their early life stages, are predicted to be among the most OA sensitive taxa, as OA interferes with the calcification process. In addition, mussels and oysters possess a relatively low ability to compensate for CO2 induced disturbances in extracellular body fluid pH with potential physiological downstream effects such as elevated metabolic maintenance costs. As mussels and oysters are key habitat forming organisms in many highly productive temperate coastal communities, negative OA effects may translate into deleterious effects at an ecosystem scale. In particular, the relative long generation time of most marine bivalves raises the concern that the rapid rate at which OA occurs may outpace species’ ability to genetically adapt, leaving pre-existing genetic variation as a potential key to species resilience under OA. Against this backdrop, this thesis contributes to the understanding of physiological mechanisms that underpin and define the OA vulnerability of ecologically and economically important mussels and oysters. Thereby, emphasis was placed on investigating intra-specific variance as a proxy for potential adaptive capacities. Kiel Fjord is located in the Western Baltic Sea and is characterised by strong seasonal and diurnal fluctuations in seawater PCO2. These fluctuations are caused by upwelling events of acidified bottom waters with peak PCO2 values (>2300 μatm) that are already by far exceeding those projected for open ocean surface waters by the end of this century. Despite these unfavourable conditions, blue mussels (Mytilus edulis) dominate the benthic community, which makes this population particularly interesting in the context of metabolic adaptation to OA. Consequently, a long-term multi-generation CO2 acclimation experiment with different family lines of M. edulis from Kiel Fjord formed the first part of this thesis. Offspring of 16 different family lines were transferred to three different PCO2 conditions, representing present and predicted PCO2 levels in Kiel Fjord (700 μatm (control), 1120 μatm (intermediate) and 2400 μatm (high)). Larval survival rates were substantially different between family lines at the highest PCO2 level. Based on these differences, families were classified as either ‘tolerant’ (i.e. successful settlement at all PCO2 levels) or ‘sensitive’ (i.e. successful settlement only at control and intermediate PCO2 level). Subsequently, the offspring were raised for over one year at respective PCO2 levels, followed by measurements of physiological parameters at the whole-animal, tissue (gill and outer mantle) and biochemical level (key metabolic enzymes). The results revealed that routine metabolic rates (RMR) and summed tissue respiration were increased in tolerant families at intermediate PCO2, indicating elevated homeostatic costs. However, this higher energy demand at the intermediate PCO2 level was not accompanied by a simultaneous increase in energy assimilation (i.e. clearance rates (CR)), indicating an incipient imbalance in energy demand and supply. Consequently, RMRs at the highest PCO2 were not different to control RMRs but associated with reduced CRs, which correlated with a lower gill metabolic scope, reduced gill mitochondrial capacities (lower capacities for citrate synthase (CS) and cytochrome c oxidase (COX)) as well as an increased capacity for anaerobic energy production (lower ratio of pyruvate kinase to phospoenolpyruvate carboxykinase). In conjunction with a lower COX to CS ratio observed in outer mantle tissue, this suggested a CO2-induced shift of metabolic pathways in tolerant families at the highest PCO2 level. By contrast, sensitive families had an unchanged RMR, tissue respiration and CR at the intermediate CO2. However, a higher control RMR in sensitive than tolerant families at similar CR suggested a lower, CO2 independent metabolic efficiency in sensitive families. This was also reflected in their lower gill mitochondrial scope at control conditions compared to tolerant families. These findings suggested that sensitive families lack the metabolic scope to cover OA induced higher maintenance costs and have to rely on energy reallocation and thus, energy trade-offs which may also have prevented survival at the highest experimental PCO2 level. Accordingly, investigations of 3-hydroxyacyl-CoA dehydrogenase (HADH) capacities, which catalyses a key step in lipid oxidation, suggested an increased reliance on lipids as metabolic fuel in sensitive families at elevated PCO2. If this was also prevalent during the larval phase, a quicker depletion of lipid reserves before completion of metamorphosis may have contributed to the higher larval mortality at the highest PCO2 treatment in sensitive compared to tolerant mussels. The second part of the thesis aimed to clarify whether a higher OA tolerance in Sydney rock oysters (Saccostrea glomerata) is directly correlated with an increased capacity to compensate for CO2 induced extracellular acid-base disturbances, and whether such a capacity is driven by higher metabolic and ion-regulatory costs at the tissue level. Earlier studies focusing on two different populations of Sydney rock oysters demonstrated that oysters that were selectively bred for increased growth and disease resistance (‘selected oysters’) have a higher CO2 resilience compared to the wild population (‘wild oysters’). To unravel the underlying physiological mechanisms, oysters of both populations were acclimated at control and elevated PCO2 (1100 μatm) levels for seven weeks, followed by determinations of extracellular acidbase parameters (pHe, PeCO2, [HCO3 -]e), tissue respiration and indirect determination of energy demands of major ion regulatory transport proteins. Indeed, at elevated PCO2, wild oysters had a lower pHe and an increased PeCO2 whereas extracellular acid-base status of selected oysters remained unaffected. However, differing pHe values between oyster types were not driven by elevated metabolic costs of major ion regulators at tissue level. Selected oysters rather exhibited an increased systemic capacity to eliminate metabolic CO2, which likely came through higher and energetically more efficient filtration rates and associated facilitation of gas exchange, suggesting that effective filtration and CO2 resilience might be positively correlated traits in oysters. In conclusion, the findings of this thesis contribute to the growing evidence that ongoing OA will likely impair the physiology of marine mussels and oysters with potentially associated downstream consequences for the respective ecosystems. However, the results also suggest adaptive capacities in both species studied. The higher CO2 resilience of selected Sydney rock oysters was expressed within the – in evolutionary terms – rapid time span of only a few generations of selective breeding, which indicates that rapid adaptation to OA may be possible in marine bivalves. The observed intra-specific variation of OA responses in blue mussels suggests standing genetic variation within this population, which is likely to be key for the persistence of populations under rapidly occurring OA. However, as global change is not limited to OA, future research will have to reassess potential resilience traits and adaptive capacities to OA when combined with changes in other environmental drivers.