Attributes analysis is a commonly used data processing and interpretation technique in seismic exploration. The purpose of this study, attribute analysis is proposed, developed and implemented for GPR data to reveal detailed information in detecting the presence of hydrocarbon contaminants. One known hydrocarbon contaminated sites were investigated in order to verify its feasibility. GPR profiles acquired from field, in conjunction with actual field sampling along with the interpretation of attributes are demonstrated in this study. Finally, a conceptual site model of LNAPL contamination is illustrated to explain the observation from attributes. This study gives a good example that thin beds caused by floating LNAPL could be a useful signature for detecting hydrocarbon contamination.