In the institutional literature, studies to formally incorporate institutional quality in spatial econometric modelling to account for its spatial dependence are undoubtedly of a recent vintage, see for example Kelejian et al. (2008), Faber and Gerritse (2009), Bosker and Garretsen (2009), Claeys and Manca (2010) and Arbia et al. (2010). These studies examine spatial spillover effect of institutional quality and uncover the channels of the institutional spatial spillover effect. Earlier, there are a number of empirical studies whose findings support the existence of institutional spatial dependence between neighbouring countries, such as Ades and Chua (1997), Easterly and Levine (1998), Murdoch and Sandler (2002) and Simmons and Elkins (2004); these studies however do not employ a formal spatial econometric framework. In this study, we employ a formal spatial econometric namely spatial lag model to investigate the spatial dependence between growth and institutions. A standard neoclassical growth model augmented with institutional controls is estimated and spatial lag model is found to be the most appropriate model to account for the spatial dependence. This indicates that the spatial externalities are not a nuisance factor, but a substantive one. Institutions are proxied by two indices reflecting property rights and political institutions and they are shown to significantly determine economic growth in developing countries and the institutional spillover is found to exist between the countries via an indirect route, i.e. institutions in a country lead to economic improvement in that country and subsequently give impact to the neighbouring countries income growth. This study is organized as follows: section 2 explains the model and data sources, while in section 3 estimations results are discussed. Section 4 concludes.