The National Program for Sustainable Irrigation provided funds to undertake an 8-week research study on-farm during the 2010-2011 summer. The initial approach to this study sought to investigate and improve irrigation application efficiency of furrow irrigated sugarcane in the Arriga Flats of Far North Queensland, using the surface irrigation model SIRMOD. Due to above average summer rainfall preventing growers from irrigating during the whole study period, an alternative approach had to be developed. Seeking to identify areas of irrigation inefficiency on a fine spatial scale over one furrow- irrigated sugarcane block, I related bulk soil electrical conductivity (ECa) measurements, obtained from an EM38 survey, to deep drainage (DD) estimates as calculated by the model SALFPREDICT, using analytical data from 8 selected soil profiles. Using simple linear regression analysis, the study found that in this area (with soils that are variably sodic but not saline), the ECa was a poor predictor of estimated DD (r2=0.12). The study also found estimated DD to vary considerably (40-254mm/year) within the block, even though it was mapped as having one soil type. Appropriate identification of areas of irrigation inefficiency and the management of areas most susceptible to DD losses therefore requires a fine scale quantification of soil properties which influence DD.