High-resolution paleoclimatic data are an essential requirement for testing numerical models of climate change and the global carbon cycle. If the long tree-ring chronologies, originally established for the purpose of dendrochronology, are to be fully exploited as an indirect measure of past climatic variability, additional techniques are required to obtain this information. The determination of the δ13C value of tree-ring cellulose has been used successfully to reconstruct past climates. However, under both aerobic and anaerobic conditions, the polysaccharide components of vascular plants (mainly cellulose and hemicelluloses) are more prone to rapid degradation than lignin. This has serious implications for the use of carbon isotope values of tree-ring cellulose as an indirect measure of past climates. An absolutely dated ring-width chronology was established for oaks (Quercus robur L.) growing at Sandringham Park in eastern England. Carbon isotope values were determined on α-cellulose and ‘Klason’ lignin isolated from annual latewood samples over the period AD 1895–1999. The carbon isotope values of earlywood lignin are correlated with the latewood carbon isotope values of the previous year, supporting the theory that some of the carbon utilised in earlywood synthesis is assimilated in the previous year. The high-frequency variance in the carbon isotope indices of latewood lignin and cellulose is highly correlated with combined July and August environmental variables, indicating that they were formed at similar times. There was no evidence of secondary lignification. These results demonstrate that the determination of carbon isotope values of latewood lignin offers the potential to obtain unambiguous proxy climatic data covering several millennia.