Aerial thermography was first applied in Iceland in the early sixties in conjunction with the suboceanic eruption of Surtsey. The same technique was subsequently applied (in the late sixties, early seventies) to some major geothermal areas in Iceland. Although it gave a global view of heat on the surface, local geologists/geophysicists were not impressed by the inroad of this space technology. According to this view, the technique lacked sensitivity, resolution, means of calibration, and gave very distorted images. Although microwave remote sensing (applied in glaciology) and processing of Landsat data (including IR) was practiced, aerial thermography lay at rest in Iceland until the mid eighties, when it was taken up again first for demonstration purposes using an instrument we rented jointly with the Geodetic Institute, but later it would be based on our own developments. Our work being application driven, due to the way this development is financed, the emphasis in the late eighties/early nineties was mainly on the near infrared range. For the last three years it has shifted however towards the far infrared range again and focused on geothermal areas, both for scientific purposes as well as for natural heat reservoirs for distance heating networks. Aerial thermography of wide area natural heat and energy distribution networks sets stringent requirements, e.g. to sensitivity and temporal, spatial as well as spectral resolutions. Efforts to meet that goal in Iceland have led to new developments based on wide aperture line scanners covering swaths in the flight direction, frame based focal plane cameras and on improved signal processing and data processing procedures. A particular emphasis has been on error corrections due to unpredictable aircraft movements.