Architrack, a Salzburg-Vienna cooperation, uses no-touch infrared reflection oculography as an evaluation technology. The US-fabricated eyetracker continuously measures the position of the left eye axis in 20 msec intervals. It was the instrumentation to record the architectural preferences of architects and of laypeople (experiment 1) as well as for objectifying the perceptual influence of urban gestalt factors by F. Moser (experiment 2). Surprisingly, eyetracker technology was only once applied to architectural stimuli in a project at Lund (Sweden) and, within IAPS, in a design evaluation by S. Lengyel (1988). The experiments presented here were realized in a diploma thesis at Salzburg University. Experiment 1 presented six Global Architecture images as wall projections by a video beamer and measured the fixation times of five professional architects and of ten laypeople (5 male, 5 female) within ten second intervals in a first run. In a second run, verbal comments on the stimuli were recorded. It was found out that the “professional focus” hypothesis (architects show significantly more fixations on architecture) was not true but that the visual preferences of the experts were more balanced, systematic, and explorative. Triangulating the fixation measurements and verbal records image-by-image, it was seen that preferences detectable by the eyetracker resulted in more explicit verbal descriptions. Also, a succession analysis of the order of fixations per image was done. Experiment 2 used ten color images containing urban gestalt factors of F. Moser, i.e. spatial elements allegedly influencing the visual contact with the environment. The factors “gate”, “building corner”, “vegetation”, “vanishing line/point” and “eaves edge” were studied. Comparing expert and non-expert eyetracking results, it was found that the gestalt factors “gate”, “vegetation” and “vanishing line/point” had explanatory power for visual preferences of both studied groups. The Moser gestalt factors, originally illustrated by drawings without urban details such as cars, persons, street lamps or ads, structured perceptional preferences and were not lost in the abundance of other objects and symbols. Thus, the pilot study of the new system Architrack proved successful to record behavioral, involuntary data allowing to check the “eye of the beholder” parallel to a verbal record. As already common in clinical, reading, traffic, advertising and software usability research, eyetracking also has the potential to enrich our knowledge on the environment-behavior interface of architectural preference.