Avocado yields are decreased by chloride toxicity and soil salinity throughout California avocado orchards but there is little information on the extent to which different rootstocks can be used to improve tree performance under saline conditions. This research is aimed at the development of a production function model that can be used to predict the impacts of irrigation water chloride content and salinity (EC) on avocado yields. The model further evaluates the effects of different soil chemical and physical properties, water chemistry, and rootstocks on the accumulation of chloride in the leaf tissue, as well as the effects of salinity and chloride on root growth. Data are being collected from 10 orchards that span the major avocado production areas from San Diego to San Luis Obispo. Our modeling approach involves the use of an artificial neural network (ANN) program that enables us to separate out very complex interactions that cannot be detected using traditional statistical procedures. When completed, this research will allow us to provide a simple, user friendly program that growers can access via the internet to predict how different water qualities will affect their yields for Hass avocado produced on each of the different main rootstocks that are now being used in California.