Radicalization is the process by which an individual becomes willing to use politically motivated violence or illegal means, or to support those who do. It can only be understood fully by longitudinal analysis. Much existing work instead tries to reconstruct the process by looking at its outcomes: those who have become radicalized. The result is a large number of theories and mechanisms, with an absence of compelling empirical support. Theories of radicalization make implicit predictions about the variations among attitudes in the communities from which radicals are drawn. We report the results of a survey of attitudes to issues widely believed to be relevant to radicalization among Ottawa Muslims in 2008. Analysis of the patterns of variation in attitudes is not consistent with popular theories of radicalization. In this population, attitudes vary along three independent dimensions: social economic political satisfaction dissatisfaction, moral religious satisfaction dissatisfaction, and a dimension that seems plausibly to be associated with radicalization. This suggests that attempts to prevent radicalization by, for example, providing better economic support, by mitigating unpopular policies, by addressing discrimination and other grievances, and even by emphasizing moderate forms of religion may not have much effect on reducing radicalization in Canada.