Abstract: Abstract: Soil moisture content is one of the key influence factors of crop growth. Identifying variation trends of soil water content and its precise simulation has important practical significance for keeping more effectively moisture, increasing crop water use efficiency, and enhancing the capacity of drought disaster prevention. Based on soil moisture observation data, temperatures, rainfall data and development period in maize growing season from 2008 to 2011 at Jinzhou, and combined with soil moisture module in CERES-MAIZE, the rain-fed maize farmland spatial and temporal variation characteristics of soil water was analyzed and simulated. The results showed that: the growing season precipitation did not reflect the variation trend of soil moisture conditions, the precipitation and precipitation frequency under moderate rain was correlated well with soil water content during growing season; Soil moisture was increasing with soil depth, the average value and lowest value of soil moisture from 0 cm to 40 cm of soil depth were 69%-82% and 49%-82% of field capacity, respectively. It was a quadratic curve relationship between Corn root biomass and the growth time. The explain rate of the model built was 89.7%. Coupling root biomass model and leaf area index model with soil moisture module can better simulate the soil moisture spatial and temporal variations of soil moisture in rain-fed maize ecosystem.