Conventional urban economic analysis suggests that a local economy's size is closely related to a number of features, including levels of human capital and the availability of specialized inputs, which are likely to influence positively the rate at which it accumulates further economic activity. At the same time, urban theory also suggests that once cities reach a certain level of size, these agglomeration benefits begin to peter out, while diseconomies rise rapidly. Consequently, we should see an inverted U-shaped pattern of growth with respect to economic size-rates of growth first rise, then fall as size increases. This paper shows that, while such a pattern is largely absent from recent data on growth in metropolitan area population and employment, it emerges strikingly in county-level data.