Abstract: Biochar is a carbon-rich solid product obtained from heating biomass under oxygen-limited conditions. Biochar application has the potential to mitigate greenhouse gas emission. Dryland farming areas in Northwest China emit substantial amounts of greenhouse gases. The aim of this study was to determine the effects of different biochar rates on diurnal variations in methane (CH₄), nitrous oxide (N₂O) and carbon dioxide (CO₂) emissions in the western Loess Plateau. Treatments included 6 biochar application rates (3 replications): 0 t·hm² (control, B0), 10 t·hm² (B1), 20 t·hm² (B2), 30 t·hm² (B3), 40 t·hm² (B4) and 50 t·hm² (B5) t·hm². Soil moisture and temperature were measured concurrently with gas measurement. The results showed distinct diurnal variations in CO2, CH₄ and N₂O fluxes for different biochar application rates. The trends of change in the fluxes of the 3 gases (CH₄, N₂O and CO₂) were consistent with daily variations in temperature. Daytime fluxes were greater than nighttime fluxes. The order of absorption peak of CH₄ was B0 (10.14ug·m^-2·h^-1) > B1 (7.82 ug·m^-2·h^-1) > B2 (6.57 ug·m^-2·h^-1) > B5 (2.89 ug·m^-2·h^-1) > B4 (1.05 ug·m^-2·h^-1) > B3 (0.10 ug·m^-2·h^-1). A similar order was noted for average emission flux of N₂O, given as B0 (288.79 ug·m^-2·h^-1) > B1 (201.78 ug·m^-2·h^-1) > B5 (164.02 ug·m^-2·h^-1) > B2 (157.14 ug·m^-2·h^-1) > B4 (154.60 ug·m^-2·h^-1) > B3 (112.06 ug·m^-2·h^-1). The order of average emission flux of CO₂ was B0 (85.44 mg·m^-2·h^-1) > B1 (80.91 mg·m^-2·h^-1) > B2 (76.49 mg·m^-2·h^-1) > B5 (69.10 mg·m^-2·h^-1) > B4 (67.19 mg·m^-2·h^-1) > B3 (65.29 mg·m^-2·h^-1). The results showed that when biochar input was less than 30 t·hm², mean emission fluxes of CH₄, N₂O and CO₂ dropped with increasing biochar application rate. However, when biochar input exceed 30 t·hm², the mean emission fluxes of CH₄, N₂O and CO₂ increased with increasing biochar application rate. The soil was a good source of atmospheric CH₄ for all treatments (except for 30 t·hm²) and sources of atmospheric N₂O and CO₂, irrespective of treatment. Soil temperature at 5 cm depth was correlated with biochar application rate — y = 0.017 6x + 16.585 (R2 = 0.302 6, r = 0.55, P < 0.05), but soil moisture at 5 cm soil depth was linearly correlated with biochar application rate — y = 0.056 5x + 13.626 (R²=0.815 1, r = 0.903, P < 0.05). The average fluxes of CH₄, N₂O and CO₂ under the control treatment were positively correlated with soil temperature of both soil surface and the 05 cm depth. The others treatments were also positively correlated with different levels of biochar. Biochar application at 30 t·hm² reduced greenhouse gas emission. The differences in both soil temperature and moisture caused by different input levels of biochar were the main reasons for the differences in CH₄, N₂O and CO₂ emissions. Correction coefficient and regression analysis of optimal measure time revealed that the optimal observation period of the three greenhouse gases was between 8 a.m. and 9 a.m.