This study analyses, through cross-section estimation methods, the influence of spatial effects in productivity (product per worker), at economic sectors level of the NUTs III of mainland Portugal, from 1995 to 1999 and from 2000 to 2005 (taking in count the data availability and the Portuguese and European context), considering the Verdoorn relationship. From the analyses of the data, by using Moran I statistics, it is stated that productivity is subject to a positive spatial autocorrelation (productivity of each of the regions develops in a similar manner to each of the neighbouring regions), above all in services. The total sectors of all regional economy present, also, indicators of being subject to positive autocorrelation in productivity. Bearing in mind the results of estimations, it can been that the effects of spatial spillovers, spatial lags (measuring spatial autocorrelation through the spatially lagged dependent variable) and spatial error (measuring spatial autocorrelation through the spatially lagged error terms), influence the Verdoorn relationship when it is applied to the economic sectors of Portuguese regions. The results obtained for the two periods are different, as expected, and are better in second period, because, essentially, the European and national public supports (Martinho, 2011).