he summer scholarship project was awarded to Tami Mills (BSc), currently enrolled in Masters in Engineering Technology Agriculture, at the University of Southern Queensland, Toowoomba. The scholarship commenced on 20th November, 2007 and was completed on 20th April 2008. Scholarships funds were administered through The Condamine Alliance.This project was in collaboration with CRC Project 2.1.02 &quote;Capturing our understanding of soil water balance and deep drainage under irrigation in models&quote;, a project funded in partnership with Cotton Catchment Communities CRC, CRDC, Condamine Alliance and QMDC (led by Dr Mark Silburn).The EM38 was easy to use and gave results that were highly correlated with soil moisture contents to a range of depths. Although multiple measurements are required to account for spatial variability, the measurements are quite easy to make. The collection of 192 ECa measurements (32 points at 3 heights in 2 modes in one location) took approximately half an hour.The data indicate that for each situation (head, mid, tail, row, ditch) 4 measurements are typically required in vertical mode to get an average that is within 5 mS/m of the true mean. In horizontal mode, 15 measurements are typically required for the same accuracy. The greater number of samples required in horizontal mode is presumably because the surface soil has more variability due to cracking and roughness. In larger areas, it will be important to consider the spatial distribution of water, salt and clay in the soil when sampling.When the soil is close to saturated with water, the accuracy of the results may be less than at other times. Because we only had calibration data from drained upper limit and drier, the accuracy is not known. However, similar calibration problems would exist regardless of the sensor or method used. Collecting data near saturation is extremely difficult because of problems such as vehicle access and physically removing soil samples from soil coring tubes.Raising the EM38 to sample to a range of depths was very effective. Good correlation was obtained for both orientations and all heights. Interestingly, at 0.4 m height, the variability of measurements in the horizontal mode was less than at the soil surface, so there appears to be considerable potential for using the EM38 in horizontal mode at 0.3 to 0.5 m height to measure the water content of shallow surface layers of soil.Temperature, time of day, size of the cotton plants, and other potential effects on the readings from the EM38 were non-significant.The EM38 monitoring highlighted how wet the soil was until late March. Some of this was due to rainfall soon after the irrigation in late January, but it appears that the irrigation was going to be somewhat earlier than necessary. The soil moisture deficit in late January was small when compared with later in the season.The data collected by the EM38 agreed to a large degree with the predictions from the HowLeaky? Model, which estimated that during two periods in the season, a significant amount of deep drainage occurred (100 mm). This is of concern both in terms of water wastage and the potential for it to contribute to the rise of salty groundwater and surface salinity.