The crystal and magnetic structure of LaTiO3 has been studied by x-ray and neutron diffraction techniques using nearly stoichiometric samples. We find a strong structural anomaly near the antiferromagnetic ordering, TN =146 K. In addition, the octahedra in LaTiO3 exhibit an intrinsic distortion which implies a splitting of the t2g-levels. Our results indicate that LaTiO3 should be considered as a Jahn-Teller system where the structural distortion and the resulting level splitting are enhanced by the magnetic ordering.LaTiO 3 has been studied already in the seventies and was thought to be a text book example of a Mottinsulator with antiferromagnetic order [1]. Ti is in its three-valent state with a single electron in the t 2g -orbitals of the 3d-shell. The titanate is hence an electron analog to the cuprates with a single hole in the 3d-shell. However, the t 2g -orbitals in the LaTiO 3 are less Jahn-Teller active and, therefore, the orbital moment may not be fully quenched in the titanate. The physics of the orbital degree of freedom has recently reattracted attention to this material [2,3].The ordered moment in LaTiO 3 amounts to 0.46 µ B which is much smaller than the value of 1µ B expected for a single electron with quenched orbital moment [4]. Quantum fluctuations can explain only about 15% reduction in the 3D-case. A straight-forward explanation could be given in terms of spin-orbit coupling, as an unquenched orbital moment would align antiparallel to the spin-moment in the titanate. However, in a recent neutron scattering experiment the magnon spin gap was observed at 3.3 meV, and it was argued that the strong interaction of an orbital moment with the crystal lattice implies a much larger value for the spin gap [2]. An orbital contribution to the ordered moment in LaTiO 3 was hence excluded. On the basis of standard theories, however, even the G-type antiferromagnetic ordering in LaTiO 3 may not be explained without a spinorbit coupling. Instead one expects ferromagnetism [5,6] related with the orbital degeneracy. Under the assumption of a specific structural distortion, Moshizuki and Imada recently presented a successful model for the antiferromagnetic order in LaTiO 3 [7]. However, there is no experimental evidence for such a distortion. The puzzling magnetic properties of LaTiO 3 led Khaliullin and Maekawa to suggest a novel theoretical description for RETiO 3 based on the idea of an orbital liquid. They were able to explain many of the magnetic characteristics of LaTiO 3 [3], but the presumed orbital fluctuations have not been observed [8]. Therefore, magnetism in LaTiO 3 still remains an open issue.We have reanalyzed the crystal and magnetic structure of LaTiO 3 by x-ray and by neutron diffraction samples with almost perfect stoichiometry. First, we find a clear structural anomaly at the Néel-ordering and, second, the shape of the octahedra in this compound is not ideal but distorted. From these observations we conclude that LaTiO 3 has to be considered as a soft Jahn-Teller system thereby explaining m...