This dissertation studies the puzzling divergence in immigration attitudes between Western and Central-Eastern Europe (CEE). While immigration attitudes in the West liberalised, they became increasingly conservative in the East. The gap in the willingness to accept at least some immigrants between the two regions drastically widened from ten percentage points in 2010 to more than thirty points in 2018. This gap emerged suddenly after decades of attitudinal convergence between the two regions. Its rapidity and timing question the conventional wisdom attributing anti-immigration attitudes in CEE to communist legacies and socialisation processes. Furthermore, it challenges some of the most prominent theories on cross-country variation in immigration attitudes - economic competition and group threat - as the deterioration of immigration attitudes coincided with a period of sustained economic growth in CEE and low immigration in contrast to Western Europe. To address this puzzle, this dissertation develops a new general theoretical framework explaining aggregate change in attitudes. It identifies five main groups of factors and spells out the concrete mechanisms through which these factors can contribute to country-level change in immigration attitudes. Subsequently, the empirical chapters systematically examine each mechanism, using data from 28 European countries. They propose several original indicators and estimation techniques and apply a variety of research designs. The employed methods include spatial analysis, generalised additive models, and estimated dependent variable models with feasible generalised least squares estimators to overcome several methodological issues and analytical limitations. The analyses draw on a range of data including inter alia cross-sectional surveys; geo-localised panel surveys; fine-grained contextual data; and several original datasets the most important of which is a survey experiment using a representative sample of almost 5,000 respondents in Western and Eastern regions of Germany. The results refute the conventional wisdom that attributes the anti-immigration sentiment in CEE to socialisation and emigration. Instead, they demonstrate that individuals’ immigration attitudes are malleable by migratory experience, spatial exposure to foreigners, and cues from political parties and fellow citizens. This dissertation links the sudden and rapid drop in immigration attitudes in most of the CEE region to the electoral strategy of mainstream parties. The results have implications for our understanding of attitude formation and change. They suggest that individuals’ political attitudes are not given in absolute terms. Rather, individuals may adopt positions relative to a reference, which, if changing, may lead to major attitudinal shifts.