This paper uses the Canadian 2006 Census and the Swedish 2006 register data to analyse the citizenship effect on the relative earnings of immigrants, using instrumental variable regression to control for citizenship acquisition. We ask: ‘Is there a citizenship effect and if any, in which country is it that we find the largest effect and for which immigrant groups?’ We add one further dimension, asking if the size of the co-immigrant population in the municipality has an effect on earnings. We find that the impact of citizenship acquisition is substantial in both Canada and Sweden. However, the place of birth of immigrants is important. In most cases, immigrant women in Sweden enjoy a higher citizenship premium than is the case for immigrant women in Canada. Amongst men the picture is more mixed. Most European groups receive a larger citizenship premium in Canada as compared to Sweden. Being in a city with more immigrants of the same background is better for earnings in Sweden than in Canada. However, being in a city with a lot of immigrants (regardless of origin) is better in Canada as compared to Sweden.