rely.py is a program implementing the method to detect independently repeated clades by comparing phylogenies as described in Li and Lecointre (2009) and adapted to incompletely overlapping datasets in Li et al. (2009). The comparison can be performed on trees obtained by any inference method (maximum parsimony, Bayesian inference, maximum likelihood). The program computes repetition indices, provides greedy summary trees for each validity domain and a nexus matrix representation of the clades weighted by their repetition indices. The additional script concatnexus.py assists the user in preparing the primary analyses, but it can also be used separately to concatenate nexus datasets.