Two basic lattice distortions permeate the structural phase diagram of oxide perovskites: antiferrodistortive (AFD) rotations and tilts of the oxygen octahedral network and polar ferroelectric modes. With some notable exceptions, these two order parameters rarely coexist in a bulk crystal, and understanding their competition is a lively area of active research. Here we demonstrate, by using the LaAlO 3 =SrTiO 3 system as a test case, that quantum confinement can be a viable tool to shift the balance between AFD and polar modes and selectively stabilize one of the two phases. By combining scanning transmission electron microscopy (STEM) and first-principles-based models, we find a crossover between a bulklike LaAlO 3 structure where AFD rotations prevail, to a strongly polar state with no AFD tilts at a thickness of approximately three unit cells; therefore, in addition to the celebrated electronic reconstruction, our work unveils a second critical thickness, related not to the electronic properties but to the structural ones. We discuss the implications of these findings, both for the specifics of the LaAlO 3 =SrTiO 3 system and for the general quest towards nanoscale control of material properties. DOI: 10.1103/PhysRevLett.119.106102 Polar lattice distortions [1,2] and "antiferrodistortive" (AFD) tilts of the BO 6 octahedral network [3,4] are ubiquitous in the physics of ABO 3 perovskite oxides and largely responsible for their interesting functional properties. Indeed, the former potentially leads to ferroelectricity, while the latter often has a dramatic impact over the orbital and magnetic orders. Although some notable exceptions exist (e.g., BiFeO 3 ), the two order parameters tend to be mutually contraindicated in a bulk perovskite crystal [3]. Such a competition has often been used in the continuing search for new multiferroics or, more generally, to engineer exotic phases via atomic-scale design, as in the recent report of polar metals at perovskite interfaces [5]. Overall, controlling the balance between the competing polar and AFD phases offers exciting opportunities to realize novel properties "à la carte." Such a balance is often delicate, as both types of lattice distortions and their mutual couplings are governed by a subtle interplay of short-range covalency (related to B-O bonding) and long-range electrostatic (or elastic) effects [6].Atomic-scale confinement appears as an ideal tool to tune the competition between different order parameters, as it is known to profoundly alter the spectrum of lattice and electronic excitations of crystalline materials [7][8][9][10][11]. In ferroelectrics, for example, size effects play a dramatic role in determining the stability of the polar state [12,13]. The driving force that determines the critical thickness for ferroelectricity is of an electrostatic nature and manifests itself as a "depolarizing" electric field that is antiparallel to the film's polarization. Therefore, the question of whether nonferroelectric (and hence insensitive to macroscopic depol...