If the baryon asymmetry of the Universe is produced by leptogenesis, CP violation is required in the lepton sector. In the seesaw extension of the standard model with three hierarchical right-handed neutrinos, we show that the baryon asymmetry is insensitive to the Pontecorvo-Maki-Nagakawa-Sakata phases: thermal leptogenesis can work for any value of the observable phases. This result was well known when there were no flavor effects in leptogenesis; we show that it remains true when flavor effects are included. DOI: 10.1103/PhysRevLett.99.161801 PACS numbers: 14.60.St, 11.30.Fs, 13.15.+g, 14.60.Pq Introduction.-CP violation is required to produce the puzzling excess of matter (baryons) over antimatter (antibaryons) observed in the Universe [1]. If this baryon asymmetry of the Universe (BAU) was made via leptogenesis [2], then CP violation in the lepton sector is needed. So any observation thereof, for instance in neutrino oscillations, would support leptogenesis by demonstrating that CP is not a symmetry of the leptons. It is interesting to explore whether a stronger statement can be made about this tantalizing link between low-energy observable CP violation and the BAU.In this Letter, we address a phenomenological question: ''Is the baryon asymmetry sensitive to the phases of the lepton mixing matrix (PMNS matrix)?'' Electroweak precision data were said to be sensitive to the top mass, meaning that a preferred range for m t could be extracted from the data. Here, we ask a similar question, assuming the baryon asymmetry is generated, via leptogenesis, from the decay of the lightest ''right-handed'' (RH) neutrino: Given the measured value of the baryon asymmetry, can an allowed range for the PMNS phases be obtained?It was shown in [3] that the BAU produced by thermal leptogenesis in the type 1 seesaw, without ''flavor effects,'' is insensitive to PMNS phases. That is, the PMNS phases can be zero while leptogenesis works, and the CP asymmetry of leptogenesis can vanish for arbitrary values of the PMNS phases. In fact, the ''unflavored'' asymmetry is controlled by phases from the RH sector only, and it would vanish were this sector CP conserving. However, it was recently realized that lepton flavor matters in leptogenesis [4 -6]: In the relevant temperature range 10 9 ! 10 12 GeV, the final baryon asymmetry depends separately on the lepton asymmetry in s, and on the lepton asymmetry in muons and electrons. So in this Letter, we revisit the question addressed in [3], but with the inclusion of flavor effects. Our analysis differs from recent discussions [7] (2RHN model), [8,9] (CP as a symmetry of the N sector), [10] (sequential N dominance) in that we wish to do a bottom-up analysis of the three generation seesaw. Ideally,