A review of the grey and published literature on field investigations of components of the water balance of irrigated pastures and annual broadacre crops in Australia culminated in the development of REFIRR, a database of studies of the water balance and irrigation management for irrigated crops and pastures. The review highlighted a lack of good quantitative data for components of the water balance across the range of crops, climatic regions, site and seasonal conditions and management. It is impossible to carry out comprehensive determinations for all but a few situations, and water balance models must be used to estimate crop water use requirement and deep drainage for the range of situations. Such models need to be evaluated against quantitative data across a range of environments. However, there are few studies that have closed the water balance and which would allow rigorous testing of water balance models. Furthermore, reported water balance studies often do not provide sufficient contextual information, nor changes over time, for adequate testing of models. Experimental studies need to place greater emphasis on providing adequate data and sufficient documentation and databases to allow future studies to use the collected data.