During the last two decades, many studies have described the evolution of the spatial distribution of immigrant groups within an urban area (for example, by using segregation indexes). Nevertheless, the factors that could explain the growth of the immigrant population within an urban area have not yet been fully explored. Moreover, the studies referred to have mainly been conducted in countries that have traditionally received large populations of immigrants, such as the USA, Canada, the UK and Northern Europe. In regard to European cities, the case of the Barcelona metropolitan area (BMA) is particularly relevant: the percentage of immigrants in the total population has increased significantly within a very short period of time (from 5.05 % in 2001 to 15.16 % in 2008, 12.03 % in 2013). Consequently, the main objective of this study is to examine the factors influencing the relative variation in the percentage of non-EU immigrants during the period of stronger growth (2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008). From a methodological point of view, we constructed two spatial models at the census tract level: the spatial lag and spatial error models. The predictors that we selected encompass several dimensions: socioeconomic status (unemployment, low education, household income and blue-collar workers), immigrant diversity (entropy), housing (small dwellings and condition of dwelling), and population density and distance to the central business district (CBD). According to the results of the spatial models, the most important factors explaining the growth of immigrant populations are, in descending order: household income, small dwellings and immigrant diversity.