Geophysical Environmental Research (GER) 63 channel scanner data acquired over Makhtesh Ramon, Negev, Israel, were investigated to retrieve information about atmospheric gases. Converting the data into apparent reflectance units via radiometric calibration, solar curve utilization and an offset factor revealed spectral features that matched modelled atmospheric gas spectra. The modelled spectra were generated using Lowtran-7 code and flight conditions. Most of the absorption features in the GER-reflectance spectra matched modelled features. Inconsistencies between the peak intensity at the oxygen bands (# 18 and #24) in the modelled and GER-reflectance spectra suggest error in the GER radiometric calibration for those bands. The modelled gas spectra indicate approximately 1-6cm precipitable water vapour for the flight and show thai the local topographic relief of 500 m can be detected from the spectral slope between 1·680-2·005/mi region (1·68/1·90 and 1·80/2·005) and from the normalized depth of the water vapour and carbon dioxide features at I-800/<m and 2005/im respectively. The correlation with the elevation is best shown by ln[A(λ)/R(λ)] at λ=l·800μm and λ — 2·005fim where A stands for the reflectance at the gas absorption wavelength A, and R for the reflectance at this wavelength if no gases occur. Both parameters (slope and \n[A(λ)/R(λ)] at around λ = 1·680-2·005μm) were calculated on a pixel-by-pixel basis and revealed good correlation with the topography. We conclude that these relations best describe the water vapour and carbon dioxide gas content as viewed by the GER sensor through the optical path length. Surface reflectance features such as clay minerals were found to enhance the slope parameter and to result in overestimation of the water vapour and the carbon dioxide contents by about 10 per cent. Although the calculations described here do not provide absolute water vapour or carbon dioxide values, these results show that quantitative detection of atmospheric gases from GER data may be possible.