Many vegetation indices have been developed to estimate crop N status at leaf and canopy levels. However, research is lacking to compare these indices for estimating plant N concentration of winter wheat (Triticum aestivum L.) at different growth stages using a common dataset. The objective of this study is to identify promising hyperspectral vegetation indices for estimating winter wheat N concentration and compare them with a series of published vegetation indices using field data from farmers' fields. Three field experiments involving different winter wheat varieties and 4 to 6 N rates were conducted with cooperative farmers during 2005-2007 in Shandong Province, China. Data from 69 farmers' fields were used to evaluate the hyperspectral vegetation indices for estimating winter wheat N concentration. Two classes of indices were tested involving all possible two band combinations from 350 nm to 1075 nm: simple ratio index (SRI) and normalized difference index (NDI). Results indicated that the performance of different hyperspectral vegetation indices and band combinations differed with growth stages and years.