Preference relations are very useful to express decision makers' preferences over alternatives in the process of decision-making. However, multiple self-confidence levels are not considered in existing preference relations. In this study, we propose a new type of preference relations: fuzzy preference relations with self-confidence. A linear programming model is proposed for estimating priority vectors of this new type of preference relations. Finally, two numerical examples are provided to demonstrate the linear programming model, and a comparative analysis is used to show the influence of self-confidence levels on the decision-making results.