Selecting participants is a key component in the design, operation, and outcomes of public-deliberation events to reflect the diversity of opinion in the larger public. This article examines how the City of Edmonton selected the members of its Citizen Panel on budget priorities in 2009 to inform Edmonton's 2010-2011 budget. The organizers relied on random selection stratified by gender, age, length of residence in the city, educational attainment, and income. This article also reports on the findings from pre-event and post-event public-opinion surveys sent to 5,000 citizens. It argues that the selection method for the Citizen Panel was justified because there were distinct views on key issues among different groups of citizens.