I propose a new test for the existence of Önancing constraints that can be applied to a variety of settings. The approach builds on tests based on investment-cash áow sensitivities and deals with two major criticisms of the standard approach. First, I use a real measure, namely Örm productivity, to control for investment opportunities instead of a proxy for the Önancial measure Q. Theoretically, Örm productivity predicts investment in an economy in which only real characteristics matter, and it is not subject to the criticisms that have been raised about Q as a control for investment opportunities. This allows me to test for the uncontroversial prediction that in the absence of Önancing constraints the reaction of investment to changes in cash áow is not di§erent from zero when investment opportunities are controlled for. In addition, productivity can be estimated using most standard Örm-level data sets while Q cannot,
which makes the suggested test more widely applicable, in particular to small Örms and to developing country settings. Second, in the empirical part, I use the Örm managersí own assessments of their di¢ culty to obtain outside Önancing as a measure of the wedge between the internal and the external cost of Önancing. I apply the test to uniquely suited Örm-level data covering 31 countries. The empirical Öndings are consistent with Önancial constraints signiÖcantly constraining Örm growth.