Assessing empirically to what extent citizens’ preferences are considered during the elaboration of public policy remains a challenge for political science research, even though it concerns one of the central pillars of modern democracies. Governments in most democracies make extensive use of public opinion research, especially in times of multiple crises. However, we do not know much about the way they mobilize this resource. While recent research examines the factors that determine the intensity of government polling at different moments of the electoral cycle and the different logics of representation behind this activity (Durovic and Schnatterer 2023), empirical evidence on the more qualitative aspect of government polling is still lacking. What types of policy issues are covered by government polls? Why are some topics overrepresented at the expense of others? Our paper sheds light on issue selection in government-commissioned opinion polls. Understanding governments as actors in the production of public opinion, not just as passive consumers, we study government polls as dependent variable. Public opinion is often considered as a possible actor in the process of the emergence of new issues. In all these models, public opinion is regarded as an exogenous entity that is known to the researcher (or at least can be measured) and, as far as we know, there is no single study on the political agenda of survey opinion. If we consider, however, that public opinion, at least in its surveyed form, is itself a social construction, it becomes necessary to focus as a first step on the political agenda of opinion polls and their agenda-setting dynamics. To do this, we develop an innovative research design and systematically analyze the factors that determine why an issue makes it onto the government's agenda. We present evidence from Germany, mobilizing an original database of all survey questions directly commissioned by the German federal government during the 18th and 19th legislative periods (2013-2021). Using a conditional logit approach, we analyze how different types of issues (regulatory, distributive and redistributive), the interest group density in the policy domain and institutional constraints in the German federal system affect the likelihood that policy issues are covered by government polls. We also control for government priorities, different types of salience (national, personal and media salience) and issue ownership by government. Our findings show that the commissioning of public opinion polls by governments takes place in a competitive environment, and that governments are torn between citizen priorities, interest group lobbying, and institutional constraints.