Coordination failures was a key issue of Robert Clower’s 1965 paper that inspired the literature on fixed-price equilibria. This literature has been the subject of a number of studies in history of macroeconomics during the last two decades (Backhouse and Boianovsky, 2013; De Vroey, 2004, 2016). Those studies generally concentrate on the 1970s and ignore the efforts made by disequilibrium theoreticians to dynamize the static framework of Robert Barro and Herschel Grossman (1971). This led historians to insist on the static nature of non-Walrasian models and on the distance between those models and the research program initiated by their precursors, Patinkin and Clower, that was resolutely dynamic. But as Plassard, Renault and Rubin (2021) showed in the cases of Bénassy and Malinvaud, the attempt to dynamize disequilibrium models was at the core of their research programs. Actually, one can consider that they proposed a way to go beyond the gaps left by Patinkin and Clower in their own contributions. My paper will be a follow up to this work with Plassard and Renault. It will raise the following question: once non-Walrasian works on dynamics are taken into account, how does it change our understanding of the fate of this research program in the 1980s? What is particularly striking is the fact that the most advanced works on non-Walrasian dynamics were contemporary to the first equilibrium business cycle models of Lucas (1975) or Kydland and Prescott (1982). This suggest that a comparison would be interesting. I will thus present and assess the non-Walrasian models of accumulation of Malinvaud (1980), d’Autume (1980, 1985) and Picard (1981, 1993) keeping in mind the characteristics of their New Classical counterparts. My claim will be that the usual arguments to explain the demise of disequilibrium macroeconomics should be completed if not reconsidered. Economists and historian of macroeconomics alike have insisted on the problem raised by the fixed-price assumption. Once dynamic models are considered, what we get is not so much fixed-prices as rules that define the way prices vary from one equilibrium to the others. These rules are somewhat arbitrary but not so different from Calvo pricing or other arbitrary assumptions used to account for price rigidity in current DSGE models. Our study will show that Malinvaud, d’Autume and Picard were faced with a different problem. This problem is the inherent complexity of the non-Walrasian framework. This problem was already present in the static context but it proved daunting and even more obvious in the field of dynamics. The lack of tractability of non-Walrasian models impeded the elaboration of satisfying microfoundations that would have taken into account agent’s intertemporal choices. Besides, the problem raised by the multiplicity of regime hindered the analysis of dynamic properties. This lack of tractability led to a lack of fruitfulness. As we will show, results stagnated and d’Autume and Picard turned to other fields, respectively growth economics and information economics.