In a conventional aircraft attitude indicator, the artificial horizon moves in the direction opposite to the pilot's hands and the movement of the aircraft. This horizon-moving format is believed to be response-effect (R-E) incompatible. An alternative format, aircraft-moving, presents movement of the aircraft symbol that is compatible with both the pilot's hand movements and the movement of the aircraft. In the present study, nonpilot participants performed an attitude tracking task with a horizon-moving or aircraft-moving display for 6 training sessions and 1 transfer session in which the display was switched to the other format. The participants performed the task equally well and showed similar rates of improvement with the 2 displays. However, a switch of display format degraded tracking performance, indicating that learned skills did not transfer between formats. The results of a secondary choice reaction task, performed concurrently during tracking, suggested that attention can be divided between visual and auditory modalities better with the horizon-moving display than with the aircraft-moving display. We argue that the horizon-moving format is R-E compatible, as far as the tracking task is concerned, and that the effects of display format on the secondary task may be due to different modes of spatial representation being used to monitor the formats.