In practice, a dominant failure mode of brittle materials, such as fiber-reinforced thermosets is the initiation and propagation of cracks under cyclic loading. In general, the damage in composites with irregular microstructure do not occur in regular patterns. There are several state-of-the-art continuum mechanical models which take the anisotropic damage for predicting the material behavior into account. Since damage generally occurs in an anisotropic way, methods that experimentally characterize the damage anisotropy are needed.Micro-Computed Tomography systems (μCT) acquire volumetric images in a non-destructive way. By combining μCT scanning and mechanical in-situ testing, detailed microstructure data of specimens under load are generated. Based on image processing methods, the damage propagation from crack initiation to fracture is analyzed. In practice, cyclic load is a common load case for many components. Consequently, fatigue and cycle load tests are essential for a comprehensive material characterization.In this contribution interrupted in-situ μCT tests with cyclic tensile load are performed on Sheet Molding Compounds (SMC), a discontinuous glass fiber-reinforced thermset composite. Image processing methods are introduced for the experimental determination of the anisotropic damage characteristic based on volumetric images. The methods enable analyzing the spatial crack orientation distribution. In order to quantify the damage anisotropy, empirical formulations of the crack density distribution and second-order crack orientation tensor are applied. The presented work in this contribution extends the results of preliminary experimental damage investigations. In addition to previous studies, the damage anisotropy is quantified and analyzed.