Abstract
Inequality follows different patterns along a high-income sector enlargement growth. Even under Kuznets' basic assumptions inequality may show a convex shape at the end of the process. Moreover, when some of the basic assumptions are removed, the inequality path may be convex along the whole process or display a reversal from inverted to uninverted U-shapes along the same process. That is why empirical studies should specify which process and which phase within it they are dealing with on testing Kuznets' hypothesis.