During a field experiment over an ice covered area of the Gulf of Bothnia in March 1975, several different types of remote sensing equipments were tested. The purpose was to investigate the capabilities of different methods for ice reconnaissance, surveillance and forecasting. This report deals exclusively with infrared thermography performed with the airborne single line scanner TEKLA in the 8-14 pm region. The thermal recordings were concentrated to a 5x5 km2 test area, which was described in great detail by ground truth measurements and photography. During 14 runs at 300-2000 m flight altitude, the thermal radiation from the sea surface was recorded on photographic film and magnetic tape. The film recordings give a general survey of the apparent temperature variations within various parts of the mapped area. By processing the tape recorded information, a more detailed analysis of some interesting objects has been accomplished. The results have been compared with the available ground truth and aerial photographs. In the thermal imagery the ice cover is reproduced in great detail. Contrary to the visual image, even new ice can readily be delineated from open water. The radiance variations are useful for a coarse mapping of relative ice thickness but unreliable for absolute thickness measurements. The isolating effect of snow results in a thermal difference between snow-covered ice and bare ice. Characteristic features, such as raftings, ridges and cracks, can be localized. Temperature profiles generated from the tape recordings may facilitate measurements of the relative radiance along interesting traces. Infrared thermography is independent of illumination but is hampered by clouds and heavy fog. The area coverage for commercially available scanners is between 2 and 3.5 times the flight altitude. For ice surveillance an altitude of 2-5 km will be sufficient to resolve details of the order o 5-10 m with a high thermal resolution.