This paper investigates the hypothesis that the spatial concentration of economic activity might impair national growth, using the experiences of Brazil and Chile. In the first stage we have used regional data to estimate country-specific models in which regional growth is conditioned by a set of controls, considering population density as a measure of concentration. The results of this stage show that population density has a negative effect on the growth of Brazilian and Chilean regions. In the second stage we have simulated the impact of the actual changes in population observed in the regions within these countries between 1980 and 2010. Using the new densities, we have recalculated the growth of the regions, generating a new national growth rate as a weighted average of the estimated growth of the regional economies. The regional deconcentration observed in Brazil produced less negative results in terms of national growth. While the Chilean increasing spatial concentration resulted in negative effects for national growth: the national GDP growth rate was reduced by 0.8% on a yearly basis. These results replicate the findings of Brulhart and Sbergami (J Urb Econ 65:48-63, 2009) for Brazil and Chile in their cross-section of countries. Working with regional data within the countries allowed determine the gains/losses in the national GDP growth rate, and to point out the effects on each region within the countries.