Biased attention is assumed to play an important role in the etiology and maintenance of depression and depressive symptoms. In this paper, we used data from a categorization task and an associated model to assess the attentional bias of people with varying levels of depressive symptoms. Attentional bias was operationalized as the parameter estimate in a prototype model of categorization. For estimation, we used a Bayesian hierarchical mixture approach. We expected to find a positive correlation between depressive symptoms and an AB for negative material and a negative correlation between depressive symptoms and a bias toward positive material. Despite good model fit, Bayesian regression analyses revealed weak or moderate evidence in favor of the null model assuming no association between attentional preferences and depressive symptoms, both for negative and positive material.