This study presents a detailed investigation into the microseismic response of a rockmass being excavated in a high stress environment. AECL's Underground Research Laboratory, Manitoba, provided a unique opportunity to non-invasively monitor a tunnel excavation with a 3D microseismic array. The 46 m long cylindrical Mine-by tunnel was excavated at 420 m depth predominantly using a non-blasting method, therefore the damage zone and crack initiation is primarily due to the effect of stress redistribution and concentration following each excavation increment. Both manual and automated source parameter processing techniques are contrasted and used on the data set of over 20000 microseismic events of magnitude, MW =-1.5 to - 4.5. The relatively homogeneous and unfractured nature of the rockmass allowed the validity of fundamental spectral models to be tested. The seismicity is found to extend 1.0 m into the roof and floor regions of the tunnel and 0.8 m ahead of the tunnel face. Spatial and temporal trends in the source parameters are used to compare the seismic response of the two rock types present along the tunnel, namely granite and granodiorite. Seismicity starts earlier and occurs over a shorter time interval in the granite. Additionally, a late second phase of seismicity is found to occur in the granodiorite with these events having a similar magnitude to those at the excavation face. These trends, due to petrofabric and geotechnical differences in the two rock types, are important for safety reasons and mine design. The excellent sensor focal sphere coverage enabled the production of well constrained source mechanism solutions using both first motions and moment tensor analysis methods, and allowed source types to be contrasted with spectral parameters. Seismicity ahead of the tunnel face is predominantly deviatoric and it is concluded that this source type is resulting from movement on face parallel tensile cracks that formed early during the tunnel excavation