The history of economic thought knows a few of heated methodological controversies (Methodenstreit), the most famous of which was a “battle” at the end of the 19th century between Austrian and New historical schools on proper methods of social sciences. Later at the turn of the 1930–1940s a dispute arose between J.M. Keynes, on the one hand, and Keynesians (J. Tinbergen, L. Klein, T. Koopmans, O. Lange, J. Marshak, etc.), on the other hand, who radically diverged in assessing prospects of a new subdiscipline – econometrics. In the literature, this episode is known as the “Keynes-Tinbergen debate”, since the discussion was centered on the study by J. Tinbergen “Statistical Testing of Business Cycle Theories” (1939), where one of the first large econometric models was offered, the parameters of which were estimated using the method of multiple correlation. Almost instantly, this model became the standard against which numerous macromodels for different countries began to be built and econometrically estimated. Keynes sharply criticized Tinbergen’s approach (and more broadly, the whole econometrics), raising a number of serious objections – both technical and epistemological ones. In his answer, Tinbergen discussed almost exclusively the former. For a long time, Keynesians unequivocally recognized Keynes as the losing party, whom they accused of technical incompetence. However, the subsequent development of econometrics turned out to be largely focused precisely on the problems raised by Keynes. More importantly, his key – epistemological – objection has remained unanswered and still retains validity.