The complete gauge-invariant set of the one-loop QED corrections to the parity-nonconserving 6s-7s amplitude in 133 Cs is evaluated to all orders in αZ using a local version of the Dirac-Hartree-Fock potential. The calculations are peformed in both length and velocity gauges for the absorbed photon. The total binding QED correction is found to be -0.27(3)%, which differs from previous evaluations of this effect. The weak charge of 133 Cs, derived using two most accurate values of the vector transition polarizability β, is QW = −72.57(46) for β = 26.957(51)a 3 B and QW = −73.09(54) for β = 27.15(11)a 3 B . The first value deviates by 1.1σ from the prediction of the Standard Model, while the second one is in perfect agreement with it. PACS numbers: 11.30.Er,31.30.Jv,32.80.Ys Investigations of parity noncoservation (PNC) effects in atomic systems play a prominent role in tests of the Standard Model (SM) and impose constraints on physics beyond it [1,2]. The 6s-7s PNC amplitude in 133 Cs [3] remains one of the most attractive subject for such investigations. The measurement of this amplitude to a 0.3% accuracy [4,5] has stimulated a reanalysis of related theoretical contributions. First, it was found [6,7,8,9] that the role of the Breit interaction had been underestimated in previous evaluations of this effect [10,11]. Then, it was pointed out [12] that the QED corrections may be comparable with the Breit corrections. The numerical evaluation of the vacuum-polarization (VP) correction [13] led to a 0.4% increase of the 6s-7s PNC amplitude in 133 Cs, which resulted in a 2.2σ deviation of the weak charge of 133 Cs from the SM prediction. This has triggered a great interest to calculations of the one-loop QED corrections to the PNC amplitude.While the VP contribution can easily be evaluated to a high accuracy within the Uehling approximation, the calculation of the self-energy (SE) contribution is a much more demanding problem (here and below we imply that the SE term embraces all one-loop vertex diagrams as well). To zeroth order in αZ, it was derived in Refs. [14,15]. This correction, whose relative value equals to −α/(2π), is commonly included in the definition of the nuclear weak charge. The αZ-dependent part of the SE correction to the PNC matrix element between s and p states was evaluated in Refs. [16,17]. These calculations, which are exact to first order in αZ and partially include higher-order binding effects, yield the correction of -0.9(1)% [16, 18] and -0.85% [17]. This restored the agreement with SM.Despite of the close agreement of the results obtained in Refs. [17,18], the status of the QED correction to PNC in 133 Cs cannot be considered as resolved until a complete αZ-dependence calculation of the SE correction to the 6s-7s transition amplitude is accomplished. The reasons for that are the following. First, in case of cesium (Z = 55) the parameter αZ ≈ 0.4 is not small and, therefore, the higher-order corrections can be significant. Second, because the calculations [16,17,18] are performed for the...