A field experiment was conducted at Al-Wahada Research Station-35 km south-east of Baghdad to investigate the salt balance in a cultivated soil using saline water. Corn was planted in a Randomized Complete Block Design (RCBD) with four replicates including two factors: first factor salinity of irrigation water with four levels: 1.2, 2.5 and 5 and 7.5 ds.m-1 (S1, S2, S3 and S4). Second factor, leaching requirement with two levels of 0.15 and 0.25 (L1 and L2). The irrigation water was added whenever 50% of the available water was depleted. After harvesting, the weight of grain and straw were measured and the relative yield was calculated, Electrical conductivity and soluble cations and anions in the soil for each 10 cm depth from 0 to 100 cm were measured.SALTMED computer model was used to predict the salt balance in the soil, soil salinity and weight of crop for different treatments. The inputs of the model were meteorological data and plant parameters as well as soil properties. The efficiency of the model was tested. It was noticed that there was a significant decrease in grain yield with the increase in salinity of irrigation water. The reductions in yield as expressed in percentage due to water salinity were 13.8%, 22.5% and 47.8% for S2 and S3 and S4 respectively. While straw yield showed a decrease of 38.5% at S4. No significant difference was observed between S1 and S2. There was an increase in the grain yield from 3.82 ton.ha-1 at L1 to 4.17 ton.ha-1 at L2. The relative grain yield was 97.7% at S1 and it decreased to 51.24% at S4. The salinity threshold values calculated from linear regression equations were 4.12 and 4.59 ds.m-1 for L1 and L2 respectively. It has been realized that S2 level saved about 11% of applied fresh water and caused a decrease of 13.68% in the relative grain yield. While S3L2 and S4L2 treatment saved 29.8% and 49.8% of applied fresh water respectively, but those treatments reduced the relative grain yield 24.7% and 49.07%. The out puts of the tested SALTMED model reflected the effect of the different treatments on the salinity through-out root zone. It was concluded that this model was able to predict the yield for all treatments. There were good agreements between simulated and observed yield values. The average of differences between these values was 10.42%.