This paper illustrates the problems that are faced by flight test, whether in FAA testing or company testing, and emphasizes that most of these problems involve human factors decisions. The tool largely relied upon is pilot experience in the way aircraft are used in their natural environment. This experience is generally classed as empirical in nature. That experience is invaluable, but alone it may not lend itself to certification conclusions that are rationally based when evaluating the pilot interaction with new or unfamiliar types of flight deck systems. Test pilots need to know how to recognize when a cockpit systems function is properly conceived and implemented and more importantly how to recognize when the function may contribute to hazardous errors in interpretation. It is suggested that existing or planned human factors related research objectives be modified to include investigation of the boundary between good and unacceptable.