Nitrification inhibitors (NIs) are used to retard the nitrification process and reduce nitrogen (N) losses. However, the effects of soil properties on NI efficacy are less clear. Moreover, the direct and indirect effects of soil property variations on NI efficiency in minimizing carbon dioxide (CO2) emissions have not been previously studied. An incubation experiment was conducted for 40 days with two treatments, N (200 mg N-urea kg-1) and N + dicyandiamide (DCD) (20 mg DCD kg-1), and a control group (without the N) to investigate the response of ammonia-oxidizing bacteria (AOB) and archaea (AOA) to DCD application and the consequences for CO2, nitrous oxide (N2O) and ammonia (NH3) emissions from six soils from the Loess Plateau with different properties. The nitrification process completed within 6-18 days for the N treatment and within 30->40 days for the N + DCD treatment. AOB increased significantly with N fertilizer application, while this effect was inhibited in soils when DCD was applied. AOA was not sensitive to N fertilizer and DCD application. The nitrification rate was positively correlated with the clay (p < 0.05) and SOM contents (p < 0.01); DCD was more effective in loam soil with low SOM and high soil pH. Soil pH significantly was decreased with N fertilizer application, while it increased when DCD was applied. Moreover, DCD application decreased CO2 emissions from soils by 22%-172%; CO2 emissions were negatively correlated with the clay and SOM contents. DCD application decreased N2O emissions in each soil by 1.0- to 94-fold compared with those after N fertilizer application. In contrast, DCD application increased NH3 release from soils by 59-278%. NH3 volatilization was negatively correlated with clay (p < 0.05) and SOM (p < 0.01) contents and positively correlated with soil pH (p < 0.01). Therefore, soil texture, SOM and soil pH have significant effects on the DCD performance, nitrification process and gaseous emissions.