The Gangotri Miocene leucogranite is composed of several laccoliths (6-7 km long, 1.5-2 km thick), which can be divided into two sets of lenses: the southern lenses, intruded in the lower part of the High Himalaya sedimentary cover, and the northern lenses, intruded in an older porphyritic biotite-beating granite. In both cases, the magmatic fabric is commonly weak with no dominant stretching direction, although a rough E-W trend is present in the northern lenses. The magnetic fabric is characterized by the fact that drastic changes in the direction, in the magnitude of the magnetic parameters, or in the fabric type (planar or linear) may take place over very short distances. Measurements of preferred orientations in thin sections indicate that the strain regime was largely dominated by a coaxial component for both sets of lenses, in agreement with the large dispersion observed in both the field and magnetic lineations. The comparison of the granite and host rock structures shows that the leucogranite emplacement dates the onset of the extensional tectonism in the High Himalaya range and is not related to a southward directed thrust event associated with the Main Central Thrust. This is exemplified by the vertical attitude of the feeder dikes that intrude the metasedimentary rocks beneath the southern lenses. The presence of these dikes indicates in turn that magma ascent occurred by fracture propagation. The spatial disposition of the southern lenses could have resulted either from the disruption of a single laccolithic intrusion by crustal scale boudinage due to a northern gravity backslide of the top of the Tibetan Slab or from the intrusion of independent laccoliths. The use of the elastic bending theory of Pollard and Johnson (1973) shows that in both hypotheses, the current laccolith sizes are compatible with a laccolithic mode of magma emplacement. However, neither the density contrast between the magma and its enclosing rocks nor the lithological boundary between the Tibetan Slab and the overlying Tibetan metasedimentary series controlled the level of magma emplacement. Rather, flat-lying collapse structures, which intersected the upward propagating magma dikes, are the most likely causes of magma arrest. Such a mechanism was favored by the schist-rich lithology of the metasedimentary host rocks. In addition, field relationships indicate that the melt supply through the dike system was a continuous, rather than pulsed, process. Existing numerical treatments on the rates of magma transport through fractures show that in such a case, the laccoliths could have been built in less than 100 years. This short time of emplacement, the small size of the laccoliths, and their peripheral disposition relative to the Badrinath granite suggest that the Gangotri lenses may represent the initial stage of pluton accretion in the High Himalaya which ultimately gave rise to a much larger massif such as the Manaslu granite.