Kaldor put forward his technical progress function as an alternative to the neoclassical aggregate production function. It is shown that Verdoorn’s law is its empirical counterpart, although allowing for increasing returns to scale. However, both may be derived from an aggregate Cobb-Douglas production function. But aggregation problems and the Cambridge capital theory controversies have shown theoretically that aggregate production functions in all probability do not exist. Moreover, the only reason that estimations of ‘aggregate production functions’ give good results is the existence of an accounting identity. This article reconsiders the technical progress function and Verdoorn’s law, especially in the light of these problems. Nevertheless, it is shown that estimates of the law do, in fact, provide insights into the growth process very similar to those of Kaldor, but viewed from another perspective.