In this article an integrated framework of agenda-setting is proposed that incorporates the two main accounts of agenda-setting: the information-processing approach by Comparative Agenda Project scholars and the preference-centred account advanced by Comparative Manifestoes Project scholars. The study claims that attention allocation is determined at the same time by preferences, information and institutions, and that attention allocation is affected by the interactions between these three factors. An empirical test is conducted that draws upon a dataset of parliamentary questions/interpellations in Belgium in the period 1993–2000. It is found that attention in parliament is indeed driven by preceding party manifestos (preferences), by available information (media coverage) and by institutional position (government or opposition party). The evidence establishes that agenda-setting is also affected by the interactions between preferences, information and institutions. Actors, given their preferences, treat information in a biased fashion, and institutions moderate information's role.