The methodology for delineating water bodies on multispectral remote sensing imagery was examined and evaluated. A supervised approach is tested with the aim to accurately detect inland water, moisturised soil surface and swampy patches on the Landsat TM 7 scene. The goal of this research is to investigate whether the application of remote sensing image interpretation could further refine the possibilities of future soil conductivity measurement research. The methodologies used were the application of supervised classification algorithms based on the training data collected in the area. The achieved overall classification accuracy value of 83.0795% suggests that the methodology could be used as a successful strategy to incorporate remote sensing data interpretation into soil conductivity measurement planning and application. The main conclusion that can be drawn is that processing of multispectral data with further refinement of the presented methodologies can led to very useful outcomes for environmental measurements.