Leaf nitrogen concentration (LNC), a good indicator of nitrogen status in crop, is of special significance to diagnose nutrient stress and guide nitrogen fertilization in fields. Due to its non-destructive and quick advantages, hyperspectral remote sensing plays a unique role in detecting LNC in crop. Many studies have reported the successes of monitoring LNC with spectral techniques for main crops such as wheat, maize and soybean, but there are few researches for barley, especially malting barley, a kind of crop very demanding for nitrogen fertilization. In the study, canopy reflectance spectra (between 350 and 1050nm) from 38 typical barley fields were measured as well as the corresponding LNC in Hailar Nongken (farming cultivate), China's Inner Mongolia Autonomous Region in July, 2010. Some existing spectral indices considered to be better candidates for evaluating LNC were tested to estimate LNC in barley. In addition, the optimal combination (OC) method was tried to extract sensitive bands responding to leaf nitrogen in barley, and expected to develop a combination model for improving the precision of LNC estimates. The results showed that a regression relationship of LNC to spectral indices NPCI and PRI could well describe the dynamic changes of LNC in barley with R² of 0.67 and 0.65, RMSEs of 0.58 and 0.59, respectively. A combined model of integrating the first-order spectral derivatives of 496,499,689,797, and 882nm based on OC method exhibited the good performance with R² of 0.82, RMSE of 0.50 for LNC in barley. The high fitting revealed an excellent agreement with ground-measured truth, and indicated that hyperspectral reflectance and OC method have a good potential for assessing nitrogen concentration in barley leaves.