Industrial raw-mineral exploitation in most cases relies on rock fragmentation. In common mining and quarrying practice, the initial fragmentation is carried out by blasting due to its economic advantages. Blast-induced fines in rock negatively affect multiple aspects of raw-mineral sustainability. The Austrian Science Fund (FWF) sponsored project P27594-N29 to investigate the cause of the fines by studying blast fragmentation through small-scale blast tests and numerical simulations. This thesis covers the experimental part of the project. Various theoretical models have been developed to describe blast fragmentation and the generation of the blast fines, focussing on crushing-shearing, usually around the blast hole, and/or dynamic branching-merging of running cracks as the main underlying mechanisms. However, published studies do not cover any experimental investigation on the link between blast-induced crack development and final crack patterns with the resulting fragmentation in rock. As the blast-driven fracturing leads to the final fragmentation, this thesis focuses on the corresponding mechanisms during the development and in the final states of blast-induced crack patterns and the resulting fragmentation, considering the two main mechanisms. The blast tests were carried out by blasting confined mortar and granite cylinders with detonating cord with 6, 12, and 20 g/m of PETN. The blast-driven dynamic cracking at the frontal end face of the cylinder, opposite to the initiation point, was filmed with a high-speed camera. The post-mortem external crack patterns were captured with digital photography. The internal crack patterns and fracture surfaces were obtained with computer tomography (CT), optical microscopy, and scanning-electron microscopy (SEM). The crack patterns were used to identify the mechanisms and count topological features (i.e., crack tips and crack intersections). The blasted cylinders were screened by sieving and laser-diffraction spectroscopy to measure the blast fragmentation. The fragmentation results were correlated with the quantified mechanisms from the crack patterns and with identified microscopic mechanisms. The crack development at the frontal end face occurs in three fracture phases: 1) initial crack emerging and propagation, 2) increasingly complex branching and merging of running cracks, and 3) spalling and inrush (spillage) of the blast fumes. The topological branching/merging features are highly correlated with the used specific charge (q) and the resulting blast-induced fines. Branching/merging indicators in the high-speed images (HSI) increase with time, usually following a bilinear function with a kink point in the second fracture phase. The crack patterns become more uniform along the axis of the blasted cylinders with the increase of q. The blast loading forms a crushed zone only around the blast hole. This zone includes a compaction zone in the mortar cylinders. The thickness of the crushed zone does not directly depend on the material of the blasted cylinder, though rather on the loading and the boundary conditions. The micrographs show that the crushed zone is not only formed by crushing-shearing, as the reviewed literature suggests, though also by microscopic variations of crack branching-merging. All observed mechanisms are related to the main mechanisms in both blasted materials. Furthermore, they represent variations of these main mechanisms at different size scales, affected by the loading conditions and the micro-structure of the blasted material. An s-n(s) description of the fragmentation data shows that the main mechanisms dominate in different fragment-size ranges, which is not directly affected by q or by the blasted material.