This article revisits the genesis of modern macroeconometrics by describing the transition from John Maynard Keynes's seminal work, The General Theory of Employment, Interest and Money to the Klein-Goldberger model, regarded as the starting point for neo-Keynesian macroeconomic modelling. The first two parts discuss the antecedents of Lawrence Klein's work. At the outset, Keynes was reluctant to accept the econometric work of Jan Tinbergen. Franco Modigliani then reworked John Hicks's inaugural model into what later became the standard version of the IS-LM model. After outlining this basic context, we study Klein's intellectual itinerary. We highlight the fact that Klein set out to give macroeconomics a rigorous microeconomic foundation but gradually moved away from this goal for operational reasons. Moreover, Klein undertook the empirical work with which his name is associated because he wanted to use factual evidence to determine who-of the classical economists or Keynes-was right. We show that Klein failed to achieve this goal, having made the mistake of confusing the natural unemployment rate and involuntary unemployment. Klein's example illustrates a broader phenomenon: the gap between intention and outcome. The fact that certain intellectual contributions end up being viewed as milestones does not necessarily mean that the intentions that guided them have been fulfilled.