Seismic, is one of the extensively exploited hydrocarbon exploration tools. It is a remote-sensing technique that takes advantage of the elastic properties of the formations. For effective hydrocarbon exploration maximum possible extraction of information from seismic data set is required which is dependent heavily on signal processing algorithms e.g spectral decomposition mechanisms. Improvements in such methods is significant as inaccurate interpretation leads to invalid well locations (dry wells/water wells) which is a great loss of money (approx 25 million USD). Singularity Spectrum analysis of the seismic data using the windowing concept is proposed and is applied on real seismic data. It is found that the energy peaks of the singularity spectral attributes namely width, correlation dimension, αpeak and asymmetry within these windows indicates hydrocarbon zones. The development of a windowed singularity spectrum on seismic data for direct hydrocarbon detection is not only an alternative method but an efficient and cost saving technology in exploration and production. Copyright © 2010 ACM.