The prevailing agricultural economics epistemology of positivistic knowledge is argued to be inappropriate for the study of agribusiness. In contrast to the traditional arenas of agricultural economics study, agribusiness research should be designed to be (1) theory-building, rather than theory disconfirming, (2) capable of examining phenomena that are not readily quantifiable nor separable from context, and (3) capable of examining phenomena for which the underlying cause-and-effect structure is unstable or not given to general theory. Methods of phenomenological knowledge are much more appropriate to agribusiness research, and these methods should be adopted by agribusiness scholars and recognized for promotion and tenure purposes.