In many applications, the pairwise constraint is a kind of weaker supervisory information which can be collected easily. The constraint propagation has been proved to be a success of exploiting such side-information. In recent years, some methods of multi-view constraint propagation have been proposed. However, the problem of reasonably fusing different views remains unaddressed. In this paper, we present a method dubbed Consensus Prior Constraint Propagation (CPCP), which can provide the prior knowledge of the robustness of each data instance and its neighborhood. With the robustness generated from the consensus information of each view, we build a unified affinity matrix as a result of the propagation. Specifically, we fuse the affinity of different views at a data instance level instead of a view level. This paper also introduces an approach to deal with the imbalance between the positive and negative constraints. The proposed method has been tested in clustering tasks on two publicly available multi-view data sets to show the superior performance.