Selenium (Se) intake has decreased substantially in the UK population since 1970s. To investigate whether Se concentration in wheat grain has changed as a result of yield improvement or environmental changes, we analyzed archived wheat grain from the Broadbalk Wheat Experiment at Rothamsted, England, which has been run continuously for over 160 years. Wheat grain and soil samples were selected from plots receiving different fertilizer or manure treatments. Grain Se concentration varied from 11 to 236 ng g(-1), with a mean and median of 44 and 32 ng g(-1), respectively. Grain samples from the unfertilized control plot had significantly higher concentrations of Se than those from fertilized or manured plots; the latter received various amounts of S and also had higher grain yield. No significant trends in grain Se concentrations were detected in the fertilized or manured plots, in spite of a dramatic increase in grain yield since the introduction of modern short-straw cultivars in the mid 1960s. In the control plot, grain samples had higher Se concentrations in the periods before 1920 or after 1970 than those during 1920-1970. This temporal pattern mirrored that of SO(2) emissions and atmospheric S deposition. Soil Se concentrations showed an increasing trend in all plots over 160 years. The results show that the Se concentration of wheat grain from the Broadbalk experiment was influenced by S inputs from fertilizers and atmospheric deposition, and that improving grain yield through plant breeding has not resulted in a significant decrease in grain Se concentration in the fertilized plots.