This paper aims to review academic background, methodological tradition, applied model and data issues of agricultural economics. Emphasis was given to the discussion of model and data base on which applied researches rely. Neither models nor data sets are complete. No reliable cook-book model exists for applied research. Methodological tradition of agricultural economics seems to show that it is more efficient for problem-solving researchers to be specialized by subject matters rather than disciplines so that they could cumulatively improve both the adequacy of model and reliability of data set by specific subject matters.