Water resources planning and management requires data collection, analysis and monitoring of critical parameters like spatio-temporal water spread information and expected volumes. Temporal fluctuations in water resources can be studied through frequent spatial inventory of surface water bodies. Application of geospatial information technologies, namely Satellite Remote Sensing and Geographical information Systems (GIS), offer immense scope for replacing/supplementing the traditional monitoring methods. The traditional classification algorithms, namely supervised, unsupervised, band thresholds, Normalized Difference Water Index (NDWI) and Modified Normalized Difference Water Index (MNDWI), require a decision-based approach for each scene, which is highly subjective. Hence, suitable automatic extraction methods need to be developed. A new knowledge-based algorithm has been developed using multi-temporal spectral information available in four bands of Advanced Wide Field Sensor (AWiFS) on board ResourceSat-1 (with spatial resolution of 56 m) namely Green (G), Red (R), Near Infrared (NIR) and Short Wave Infrared (SWIR) for inventorying and monitoring of various types water bodies. The algorithm has been applied for the data obtained from other space-borne sensors with similar spectral bands such as Landsat ETM, IRS LISS III and ASTER and found to be working satisfactorily. Results were validated by comparing the results reported from other popular methods. The study provides a quick method for generation of spatio-temporal water bodies information. This will be helpful for development of Water bodies Information System (WIS) on national/global scales.