Bacillus subtilis has great effects on soil water and salt movement. In this paper, 5 different contents of Bacillus subtilis (0, 1, 3, 5 and 7 g/kg) were set, aiming to study the soil water cumulative infiltration, infiltration rate, wetting front, soil water contents and salt contents in each profile, as well as water-stable aggregate contents. The results showed that with the applying of Bacillus subtilis, the cumulative infiltration and infiltration rate in saline-alkali soil decreased obviously, compared to that in the control of soil columns, due to the water retention properties of Bacillus subtilis, especially after the initial 300 min. Moreover, with the increase of Bacillus subtilis content, the cumulative infiltration and infiltration rate first decreased and then increased. When the Bacillus subtilis content was 3 g/kg, the cumulative infiltration and infiltration rate of soil water tended to be the smallest in a certain time, and the variations between the treatments were significant (P<0.01). At the end of the infiltration, the cumulative infiltration with the Bacillus subtilis content of 1, 3, 5 and 7 g/kg was decreased by 18.49%, 21.85%, 12.18% and 3.78% respectively compared with that without Bacillus subtilis. With the applying of Bacillus subtilis, the wetting front decreased as well. At the initial time, there were no significant differences between different treatments. With the infiltration time increasing, the effect of Bacillus subtilis on wetting front in each treatment was significant (P<0.01). The wetting front took 3 700, 7 260, 7 700, 5 100 and 4 320 min to reached the bottom of the soil column with Bacillus subtilis content of 0, 1, 3, 5 and 7 g/kg, respectively, and the migration rates were 7.84×10-3, 4.00×10-3, 3.77×10-3, 5.69×10-3, and 6.71×10-3 cm/min, accordingly. When the content was 3 g/kg, the migration rate of wetting front was the smallest. Both Philip and Green-Ampt model could fit soil infiltration process well with different Bacillus subtilis contents. For Green-Ampt infiltration model, with the increase of Bacillus subtilis contents, soil water saturated hydraulic conductivity first decreased and then increased, and the wetting front suction showed an opposite trend compared with control treatment (P<0.05). For the Philip infiltration model, with the increase of Bacillus subtilis contents, the soil sorptivity and stable infiltration rate first decreased and then increased, which were obviously less than that without Bacillus subtilis (P<0.05). When the Bacillus subtilis was 3 g/kg, the minimum values of soil sorptivity, soil water saturated hydraulic conductivity and stable infiltration rate were obtained, and the maximum value of wetting front suction was obtained, which further confirmed that Bacillus subtilis could increase soil water holding capacity. When the Bacillus subtilis contents were 1, 3, 5 and 7 g/kg, the water contents were greater than that in control. At the depth of 27 cm, the water content in the treatments with Bacillus subtilis of 1, 3, 5 and 7 g/kg increased by 17.65%, 31.76%, 11.76% and 7.06% respectively, compared with that in control. Soil salt contents were also significantly decreased with the application of Bacillus subtilis. The salt content in the treatments with Bacillus subtilis of 1, 3, 5 and 7 g/kg decreased by 22.37%, 31.29%, 17.78% and 10.67% respectively, compared to that in control. After applying Bacillus subtilis, soil water-stable aggregate also increased. In the whole soil layer, the water-stable aggregates in the treatment with Bacillus subtilis of 1, 3, 5 and 7 g/kg were greater than those in control, and increased by 13.02%, 17.59%, 9.68% and 5.24%, respectively. This study can provide a theoretical support for the microorganism application in saline-alkali soil amendment. [ABSTRACT FROM AUTHOR]