Immune repertoires of T or B cells are very often studied by Complementary Determining Region 3 (CDR3) spectratyping. However, data obtained with this method is usually subject to a biased eye analysis. We developed recently the ISEApeaks software package to retrieve and handle peak data from automated sequencers, from which CDR3 spectratype data is obtained. We describe a general strategy for CDR3 spectratype analysis based on two new specific modules and multivariate statistics. The first module addresses the crucial problem of peak smoothing. The second is a toolbox for the analysis of CDR3 spectratypes, which includes perturbation computation, recurrent peak finding, expansion assessment and datamining. To illustrate our approach, we assessed the complex TCRB repertoire modifications induced by Plasmodium berghei ANKA (PbA) infection. This global and exhaustive repertoire analysis approach is of general interest for T- and B-lymphocyte repertoire studies and is currently used in human cohorts in various pathologies and during clinical trials.