Two methods of estimating reforestation area for salinity control are described. The first method, a simple water balance model, has readily determinable parameters and sensitivity. Further parameter measurements are necessary to improve its reliability and applicability. The second method is based on regressions of observed water table reductions below reforestation on the area and density of reforestation. The regressions have high correlation coefficients and are statistically significant at the 0.1% level. A groundwater modelling approach to determine the optimum reforestation distribution was found to have extensive data requirements and involved assuming values for some key model parameters. Predictions should be checked in the field. The best current approach to reforestation design is to use the water balance model or experimental regressions to predict reforestation area, and then use the groundwater model to design the optimum reforestation layout.