This paper identifies the determinants of per capita gross value added (GVA) growth in Ecuador during the 2007–2015 period, using a spatial extension of the Mankiw, Romer and Weil (MRW) model. Because as a country Ecuador is characterized by deep territorial socioeconomic imbalances, estimates using classical techniques that measure average or “global” effects would not be as justifiable and would have limited political implications. Accordingly, this study uses a spatial filtering technique, which is a recent evolution of geographically weighted regression (GWR), to account for the spatial heterogeneity of the coefficients of a growth regression that explicitly considers both physical and human capital. The results show that Ecuadorian cantons have a wide range of convergence rates and that the effect of physical and human capital varies across space.