Despite the increasing number of scientific papers on the subject, conservation planning has failed to actively prioritise the creation of new protected areas. Strategies proposed to increase the creation of new protected areas based on conservation plans include broad stakeholder participation in decision-making processes in order to include their concerns and facilitate social acceptance of proposed actions. However, there are controversial views about the effectiveness of stakeholder participation. The quality of the decision depends both on the information used by stakeholders and on how it is used, so it is necessary to evaluate both these aspects of a decision-making process. Threats are intrinsically related to conservation decisions because they are more easily understood by people than biodiversity values, and they can affect both decisions and outcomes of conservation actions. This article analyses how information about threats was used in the decision-making process conducted by the Brazilian Government in 2006 to indicate priority areas for the conservation of the Amazon biome. We first verified the consistency of the information on threats attributed by stakeholders to these new priority areas, and then assessed whether the existence, levels and types of threats influenced the choice of areas for conservation. The results showed that there were some successes in recognising threats, but also many inconsistencies, especially in assigning levels of intensity for some types of threats such as fishing. The decision-making process also did not fully use available information to indicate areas for conservation. The lack of understanding on the motivation behind these inconsistencies could suggest the presence of political opportunism. A more quantitative approach to assigning priorities is needed: one that is less dependent on the individual input of stakeholders and more accurately reflects the actual emergency status of proposed areas. This indicates that greater effort should be allocated to combining a participatory approach with a robust decision support system.