In many European countries, a growing share of population with immigrant background coincides with the surge in support for radical right parties. In this paper we show how such increases affect radical right candidacy. We use Swedish register data which identifies political candidates. With geocoded data, we match individuals running for the Sweden Democrats to their local neighbourhood contexts, and measure changes in the share of visible minority residents at scales ranging from 100 meters to 2 kilometres. For those who stayed in the same neighbourhood between 2006 and 2010, the change in the share of visible minorities generally does not affect the decision to join the pool of party candidates. This result is robust when we introduce additional tests and select on the scale of the neighbourhood, unemployment terciles, change in share of visible minority groups terciles, and entry threshold into the pool of candidates. For those who stayed in the same neighbourhood, the only significant finding is a small mobilisation effect for a subsample of individuals who live in densely populated metropolitan neighbourhoods - here we also observe a halo effect, with negative association for small-scale changes and positive association for changes in the larger halo zone.