Most of the topics and applications discussed in this book deal with ductile behavior of ice resulting in continuous deformation or creep under comparatively low stresses. This is the primary mode of glacier motion and most relevant to glacier modeling. However, most people traveling on glacier surfaces will be struck by physical manifestations of the brittle nature of ice, namely, crevasses. Crevasses are imposing chasms cutting through the surface of most glaciers that form primarily under tension when stretching cannot be accommodated by ductile flow and, instead, the ice behaves as a brittle material susceptible to rupturing. Most crevasses are confined to the upper few tens of meters and do not importantly affect glacier flow. There are, however, several reasons why crevasse initiation and propagation has received some attention in the glaciological literature. First, the orientation of crevasses-usually perpendicular to the direction of principal tensile stress-may contain information about the regional stress field (c.f. Van der Veen, 1999a). Second, fracture propagation is the primary mechanism by which icebergs break off from glacier termini, and better understanding and quantification of this process may aid in formulation a calving relation for numerical models (Van der Veen, 2002; Benn et al., 2007a,b). Third, rapid downward propagation of water-filled crevasses may be a mechanism by which meltwater ponds formed at the glacier surface can almost instantaneously penetrate the full ice thickness, thus providing a drainage route for surface water to reach the glacier bed and provide additional lubrication (Van der Veen, 2007; Das et al., 2008). For these reasons, it is worthwhile to present a discussion of crevasses on glaciers based on the review of Van der Veen (1999a). The following sections in this chapter introduce the mathematical tools for modeling crevasses and fracture propagation on glaciers with application to iceberg calving.