Using a technique that is not subject to first-order SU (3) symmetry breaking effects, we determine the Vus element of the CKM matrix from data on semileptonic hyperon decays. We obtain Vus = 0.2250 (27), where the quoted uncertainty is purely experimental. This value is of similar experimental precision to the one derived from K l3 , but it is higher and thus in better agreement with the unitarity requirement, |V ud | 2 + |Vus| 2 + |V ub | 2 = 1. An overall fit including the axial contributions, and neglecting SU (3) breaking corrections, yields F + D = 1.2670 ± 0.0035 and F − D = −0.341 ± 0.016 with χ 2 = 2.96/3 d.f. The determination of the elements of the CabibboKobayashi-Maskawa (CKM) matrix [1, 2] is one of the main ingredients for evaluating the solidity of the standard model of elementary particles. This is a vast subject which has seen important progress with the determination [3,4] of ǫ ′ /ǫ and the observation [5,6] of CP violation in B decays.While a lot of attention has recently been justly devoted to the higher mass sector of the CKM matrix, it is the low mass sector, in particular V ud and V us where the highest precision can be attained. The most sensitive test of the unitarity of the CKM matrix is provided by the relation |V ud | 2 + |V us | 2 + |V ub | 2 = 1 − ∆. Clearly the unitarity condition is ∆ = 0. The |V ub | 2 contribution [7] is negligible ( 10 −5 ) at the current level of precision. The value V ud = 0.9740 ± 0.0005 is obtained from superallowed pure Fermi nuclear decays [8]. In combination with V us = 0.2196 ± 0.0023, derived from K e3 decay [9,10], this yields ∆ = 0.0032 ± 0.0014. On its face, this represents a 2.3 standard deviation departure from unitarity [8].In this communication we reconsider the contribution that the hyperon beta decays can give to the determination of V us . The conventional analysis of hyperon beta decay in terms of the parameters F, D and V us is marred * To be published in Physical Review Letters.by the expectation of first order SU (3) breaking effects in the axial-vector contribution. The situation is only made worse if one introduces adjustable SU (3) breaking parameters as this increases the number of degrees of freedom and degrades the precision. If on the contrary, as we do here, one focuses the analysis on the vector form factors, treating the rates and g 1 /f 1 [11] as the basic experimental data, one has directly access to the f 1 form factor for each decay, and this in turn allows for a redundant determination of V us . The consistency of the values of V us determined from the different decays is a first confirmation of the overall consistency of the model. A more detailed version of this work will be published in the Annual Reviews of Nuclear and Particle Sciences [12].In 1964 Ademollo and Gatto proved [13] that there is no first-order correction to the vector form factor, ∆ 1 f 1 (0) = 0. This is an important result: since experiments can measure V us f 1 (0), knowing the value of f 1 (0) in ∆S = 1 decays is essential for determining V us .The A...