During the September--October 1989 eruption at Mt. Etna, structural (faults, folds, fractures) and seismological data (locations, focal mechanisms, shear wave splitting) on the opening of a 6.5 km long NNW-SSE fracture system were collected. The fracture system reflected the surface deformation associated to the emplacement of a blade-like dyke at shallow depth. Seismic activity occurred only at the southern tip of the fracture system and at about 1--1.5 km depth, i.e., at the interface between the less stiff clay basement and the stiffer volcanic pile. Structural data indicate that fractures formed in response to an ENE-WSW striking &sgr;3=&sgr;hmin. In contrast, the NNW-SSE compressive structures, which developed only at the southern tip of the main fracture system, were related to an ENE-WSW &sgr;Hmax. The inversion of focal mechanisms shows a thrust-type deformation related to an ENE-WSW striking &sgr;1=&sgr;Hmax. The association of compressive structures is consistent with experimental models of dykes stopping at a less stiff-stiffer interface. S-wave splitting parameters have been measured during the fracture system propagation. No change was observed in the qS1 polarization eigendirections (NE-SW), while the TD values seemed to vary with time. The results of seismological and structural analyses suggest that the dyke stopping mechanism produces: i) a seismic swarm associated to a local compressive stress field; and (ii) dyke-induced stress generating shallow aligned secondary fractures (EDA cracks). ¿ 1998 American Geophysical Union