The S-192 conical multispectral scanner, which acquired data during the 1973-1974 Skylab mission, represents one of the best sources of high resolution medium and long wave infrared data obtained by a space based sensor. The 2.1-2.35 pm and 10.2-12.5 pm spectral bands are of considerable interest since they represent spectral bands or resolutions not available from LANDSAT, DMSP, or HCMM. In this study, four S-192 data sets were investigated from the viewpoint of the statistical properties of the scenes themselves. These S-192 data sets contained scene data not normally acquired by aircraft-based remote sensors, but of interest to many applications. Represented in the data sets were: several cloud types, coastal mountains, snow covered mountains and valleys, and ocean backgrounds. The data sets were studied by choosing areas of homogeneous background and obtaining the power spectral density (PSD). The statistics of like areas were compared. Particularly of interest was the effect of clouds, both optically thin and thick, on the statistics of terrain backgrounds. In this paper, PSDs based on Skylab S-192 will be presented and infrared and statistical measures for specific background classes will be discussed.