Peak activities of radiocaesium (137Cs) in lake sediments have frequently been used to infer the ages of sediments deposited in the 1960s (137Cs derived from nuclear bomb testing) or in 1986 (Chernobyl derived 137Cs). Records of the vertical distribution of 137Cs in sediments can thus be used to provide accurate dates for a critical period in which palaeoecological reconstructions often overlap contemporary monitoring data. However, knowledge regarding how the distribution of 137Cs in sediments is affected by post-depositional processes is limited to interpretations based on the 137Cs distribution in sediments sampled at a single given date. This study assesses the extent to which the 137Cs record in annually laminated (varved) lake sediments is affected by post-depositional diffusion, using 11 archived sediment cores sampled between 1986 and 2007. The sediment record reveals how Chernobyl 137Cs incorporated into the 1986 varve diffused downwards in the core at a decreasing rate over time, whereas the surface sediments continued to receive inputs of 137Cs mobilized from the catchment soils or lake margin. In spite of these processes, all cores post-dating the Chernobyl accident had a clear and well-resolved peak in the 1986 varve, justifying the use of this feature as a fixed chronostratigraphic feature. Because of the very high levels of Chernobyl fallout at this site, downwards migration of Chernobyl 137Cs has, however, completely masked the nuclear weapons 137Cs fallout peak that had been clearly preserved in the 1964 varve of a pre-Chernobyl core sampled just three weeks before the Chernobyl accident. In consequence, the weapons fallout marker is likely to be of little use for determining 137Cs dates in areas strongly affected by high levels of Chernobyl fallout.