The North Atlantic region is a key component of the global climate system. However, the lack of long-term observations hampers our understanding of the natural variability of this region. In recent decades, bivalve sclerochronology has emerged as a field providing high-quality proxy data that can help fill these gaps and reflect marine variability at different timescales. This thesis presents centennial-scale, annually resolved sclerochronological records of hydrographic and climatic variability at St Kilda, a volcanic archipelago on the western Scottish shelf. Two floating chronologies were constructed using fossil shells of the bivalve Glycymeris glycymeris and dated to the 4th millennium before present (BP). In addition, a modern chronology was built with live- and dead-collected Arctica islandica, spanning 1889–2015 CE. Sub-annual δ18Oc data were obtained from fossil and live G. glycymeris and from modern A. islandica as a proxy for sea surface temperatures (SSTs). The growth season of both G. glycymeris and A. islandica at this location lasts from May to October, with most growth occurring before the temperature peak in August. Thus, the floating chronologies and the modern Arctica chronology represent late-spring and summer SSTs. The annual temperature range was 4.4 °C in the fossil shells, which is similar to the range observed today (3.8 °C). Average SSTs reconstructed from the fossil shells were similar to the late 19th century CE and 1 °C cooler than today. The similarity in growth season and temperature range between the fossil and modern specimens can be attributed to similar boundary conditions in the fourth millennium BP compared to today. In the final results chapter of this thesis, the modern Arctica growth chronology and an associated annually resolved δ18Oc series were compared to biological and environmental data and climate indices. Significant correlations were found between Arctica growth variability and changes in primary production, regional SSTs, zonal wind pressure, and sea level pressure. This work thus establishes that St Kilda is appropriately sited to study large-scale marine climatic variability using bivalve sclerochronology. Moreover, the Arctica chronology was positively correlated with variability in population size of the Soay sheep living on Hirta, St Kilda, which demonstrates the tight coupling of the marine and terrestrial environments at this remote location.