This paper provides a thorough test of important contextual explanations of variation in electoral support for radical right-wing parties. It has been proposed that support of the radical right is particularly strong in areas that are socioeconomically poor and/or where the concentration of immigrants is high. A variant of the latter hypothesis, known as the halo effect', states that the propensity to vote for the radical right is highest in areas close to immigrant-dense areas, but not within these areas. The data analyses are based on the total population of voting districts in Sweden (N = 5,668), which makes it possible to avoid some of the problems that usually plague studies of contextual effects on voting, such as low numbers of observations. The results demonstrate support for the socioeconomic marginalization hypothesis and, when controlling for socioeconomic factors, the halo effect hypothesis; whereas the support for group threat theory is mixed.