It is undeniable that more and more simple and rapid techniques are necessary to satisfy the increasing need of the gas detection in unconventional natural gas industry. Terahertz (THz) technique, a recently developed spectral method, was used to realize the identification and determination of the principal gas components distilled from oil shale, including methane (CH4), ethane (C2H6), carbon monoxide (CO) and carbon dioxide (CO2). An obvious variation can be observed between the components after the measurement of THz. C2H6 represented the largest refraction property because of the highest time delay relative to the reference, followed by CO2, CH4 and CO in turn. In addition, back propagation artificial neural network was employed to quantitatively characterize the components' content and the total pressure of the four-component system. The correlation coefficient of the prediction set in the quantitative model equalled 0.9955. The results indicated that THz technique combined with mathematical method is an effective tool for gas detection and can be popularized in unconventional natural gas industry. (C) 2015 Elsevier Ltd. All rights reserved.