Since its development in 1985 by Melvin and colleagues, the Attitudes Toward Prisoners Scale (ATP) has been employed to measure attitudes toward offenders. However, our current knowledge of public attitudes toward offenders is limited primarily to (a) Anglo-Saxon countries and Hong Kong; and (b) non-random samples. This paper differs from most previous research in its use of a random sample in a hitherto unexplored jurisdiction and multivariate analysis to identify correlates of public attitudes toward offenders. We identified a unidimensional structure of ATP in the context of Azerbaijan, which conforms to the findings reported by the research in the American, European settings, but contradicts the ones from Asian settings. The results from a sample of urban citizens in the capital city of Azerbaijan (n = 390) suggest that those who know a prisoner hold a more positive attitude. Demographically, those with a higher age level and holding a higher education diploma scored higher on the ATP scale. The theoretical and practical implications are discussed toward the end.