In this paper, we present a novel fracture diagnostic method to determine the geometry of multiple propagating fractures. The method relies on the measurement of the Azimuthally Resolved WEllbore Strain Tensor (ARWEST) as a function of time at multiple locations in an observation well. A pad-scale fracturing simulator is used to simulate dynamic fracture propagation in a treatment well. The geometry of the monitoring wellbore is represented with a very fine (millimeter scale) computation mesh to capture the impact of the propagating fractures on the monitoring wellbore. The axial and radial strain at different locations along the wellbore is computed as a function of time as the fractures approach the observation wellbore. These measurements together with the wellbore pressure response are interpreted to obtain the height, length and width of the fractures as well as the cluster efficiency of the stage. The emergence of peaks in the strain and pressure monitoring data clearly detects the arrival of each fracture. As the fracture approaches the monitoring well, the tensile strain measured within the wellbore in the axial direction increases, the compressive strain in the radial direction increases and the sealed wellbore pressure increases. As the fracture intersects the wellbore, the tensile strain in axial direction decreases and compressive strain in the radial direction decreases. The sealed wellbore pressure further increases. When the treatment is complete, both the magnitude of the monitored strain and pressure decrease. The major axis of the oval wellbore is oriented towards the tip of the propagating fracture. The wellbore ovality, therefore, provides a direct measure of the location of the fracture tip in 3-D. The results obtained from these azimuthal wellbore measurements can be interpreted with the aid of the simulations to provide a new low cost facture diagnostic method. This new 3-D fracture diagnostics method allows us to infer (a) the location of the fracture front, (b) estimate the geometry (length, height, width) and (c) determine the cluster efficiency by monitoring the strain tensor as a function of time along an observation well. The results presented here will allow operators to integrate the measured casing strain tensor and the sealed wellbore pressure data. Such a diagnostic method opens the possibility of real-time fracture diagnostics and optimization.