This paper reconsiders the long-run demand for M2 based on a newly constructed dataset featuring 32 countries since the first half of the 19th century. The evidence from cointegration tests suggests that a long-run equilibrium relationship for M2 demand is hardly present. Specifically, only forfive countries (Finland, Korea, Mexico, Paraguay and Taiwan) cointegration tests produce strong evidence in favor of a stable long-run money demand. Evidence for Israel and Lebanon is weaker, but still points towards a stable long-run demand for M2. For all other countries evidence speaks against a stable money demand or it is mixed across money demand specifications and/or type of cointegration test.