[en] There are numerous reasons to think that the Standard Model of physics is not the ultimate theory of nature on very small scales. However, attempts to construct theories that go beyond the Standard Model generically lead to high rates of flavour changing neutral processes that are in conflict with experiment: Quarks are the fundamental constituents of protons and neutrons. Together with electrons they form the visible matter of the universe1. They come in three generations or ''flavours''. In interactions, quarks of different generations can mix, i.e. a quark of one flavour can transform into a quark of another flavour. In the Standard Model, at first order in perturbation theory, such processes occur only via the exchange of a charged particle. Flavour changing neutral processes can only arise in processes involving loops of charged particles. This is due to the fact that all couplings of two quarks to a neutral particle are diagonal in the basis of the mass eigenstates of the quarks. There is thus no mixing of quarks of different flavour at first order. Since the loop processes are suppressed by a loop factor, the Standard Model predicts very low rates for neutral processes that change the flavour of quarks. So far, this is in agreement with experiment. In extensions of the Standard Model, new couplings to the quarks are usually introduced. In general there is no reason why the new coupling matrices should be diagonal in the mass basis of the quarks. These models therefore predict high rates for processes that mix quarks of different flavour. Extensions of the Standard Model must therefore have a non-trivial flavour structure. A possibility to avoid flavour violation is to assume that the new couplings are aligned with the mass matrices of the quarks, i.e. diagonal in the same basis. This alignment could be due to a flavour symmetry. In this thesis, two extensions of the Standard Model with alignment are studied. The first is a simple extension of the Standard Model where a second Higgs doublet is added. In such models, there are two Yukawa matrices for each fermion type. Going to the mass basis, one of them is diagonalized and together with the vacuum expectation value of the Higgs forms the mass matrix of the quarks. The other Yukawa matrix however is not diagonal. It couples two quarks and one of the mass eigenstates of the two Higgs doublets. Flavour violating processes can thus occur via the exchange of a neutral scalar. If the two Yukawa matrices were aligned for some reason this would not happen. However, the alignment can only be imposed at one energy scale and will be spoiled when evolving the couplings down to a lower scale. It is shown that in spite of this effect, alignment of the Yukawa couplings provides sufficient protection from flavour changing neutral currents to be in agreement with present experimental bounds. Another, more ambitious, extension of the Standard Model are warped extra dimensions. In these models spacetime consists of a slice of five-dimensional Anti-de Sitter space (the ''bulk'') sandwiched in between two flat four-dimensional boundaries (the ''branes''). The Higgs is assumed to live on one of the branes while all other particles are allowed to spread into the bulk. Particles that propagate in the bulk have a ''KK tower'' of heavier particles associated with them in the effective four-dimensional theory. In the bulk fermions have a vector-like mass term in addition to their Yukawa couplings to the Higgs. Via different localizations of the quarks' wave functions in the bulk, the huge differences in their masses can be explained. However, since the wave function profiles of the quarks are non-universal for the different flavours, so are the couplings to the KK excitations of gauge bosons. Rotating to the mass basis therefore introduces off-diagonal elements in these couplings and thus flavour changing neutral processes. Since the wave function profiles are a function of the eigenvalues of the vector-like masses, aligning these with the Yukawa couplings will suppress flavour violation. In this thesis a model that makes use of such an alignment mechanism is presented and is shown to be in agreement with experimental constraints.