Crop samples harvested and stored from three long-term agricultural experiments started in the 1840-1850s at Rothamsted Experimental Station (UK) have been analysed recently for Cd. Increased Cd burden in the soils of the experiments, which have had a range of treatments, originates mainly from atmospheric deposition. Soils treated with farmyard manure (FYM) or, in some cases, applications of phosphate fertilisers, have increased Cd levels. Herbage, wheat and barley grain from the three experiments were analysed by neutron activation analysis (NAA) and graphite furnace atomic absorption spectrometry (GFAAS). Samples were bulked for groups of years between 1860 and 1986, from variously treated plots in each experiment (control or 'nil' treatment, P-fertilised, FYM-amended, NPK-fertilised-limed and unlimed). There were marked differences in Cd concentrations between treatments. For example, uptake of Cd into herbage was greater where P fertiliser had been applied than not, and was greater from unlimed than limed soils. Offtake of Cd (mg ha^-1 year^-1) was affected by large differences in yield and probably also by other factors. These include changes in botanical composition in the permanent grassland experiment; cultivar changes in the wheat and barley experiments; changes in soil organic matter and soil pH of some plots; changes in atmospheric deposition of Cd through time. All of these potentially confounding factors make the interpretation of results complicated. It is concluded, however, that, with one exception, there is little evidence of a long-term increase in crop Cd concentrations at Rothamsted.