As a conventional method to determine available nitrogen in dryland soils，the modified Conway’s diffusion method should establish and optimize the conditions of alkali hydrolysis and reduction to ensure reliability of determination results. Nitrate reducing capacity and practical applicability of three common-used FeSO₄ reduction systems were compared through soil sample tests and recovery tests，aimed to provide reference for available nitrogen determination methods in the of soil agrochemical analysis. The measured results of available nitrogen in 11 soil samples indicated that 0.2 g-Fe system （0.20 g FeSO₄·7H₂O and 1.14 mol/L NaOH）was the lowest，1 g-Fe system（1.00 g FeSO₄·7H₂O and 1.68 mol/L NaOH） was middle and 1 g-ZnFe system（0.17 g Zn，0.83 g FeSO₄·7H₂O and 1.55 mol/L NaOH）was the highest. Compared with CK（no reductant and with 1.00 mol/L NaOH），0.2 g-Fe，1 g-Fe and 1 g-ZnFe systems increased the measured results by 6.9-14.0，8.9-77.6 and 10.2-169.5 mg/kg，respectively. The reduction rate of nitrate nitrogen declined with the increase of spiked level and gave a good power function relationship. Based on a recovery rate of 90%，0.2 g-Fe，1 g-Fe and 1 g-ZnFe systems respectively reduced 26.6，57.8 and 203.5μg NO₃ ^- -N in the blank spike test，while 29.0，50.3 and 153.2μg in the matrix spike test，respectively. For lower reduction rate in the actual soil sample measurement，these values were likely to be overestimating. With higher efficiency of NO₃ ^- -N reduction and wider range of application than 0.2 g-Fe and 1 g-Fe，1 g-ZnFe was more suitable for current farmland soils. Nevertheless，when a large amount of nitrate accumulated in soils，all of the three FeSO₄ reduction systems were not applicable. For alkaline-hydrolyzed nitrogen has a different significance from NO₃ ^- -N in soil and fertilizer science，it was recommended to separately measure alkaline-hydrolyzed nitrogen and NO₃ ^- -N in soil and further to integrate both as indicators of soil nitrogen supply capacity.