There is an expectation that studies examining new firm growth should take potential industry effects into account by controlling for industry. For example, Baum, Locke and Smith (2001), examined the effect of industry characteristics in terms of dynamism, complexity, and munificence. Their finding that industry mattered, but to a smaller extent than strategic choices or motivations of management, adds to the debate whether outcomes of new ventures mainly depend on environmental conditions or on the strategic choices made by the new venture’s management. However, in the new venture literature there is no consensus with regards to how industry controls should be applied, and empirical results may be contingent on researcher’s conceptualization of industry. We therefore compare the three most commonly employed measures of industry (i.e. ISIC codes converted into a singular control variable (at varying levels of definition), multi-variable measures on the basis of census data and ISIC codes, and perceptual measures of the task environment) with respect to their ability to explain new venture growth.