Steel structures have an ubiquitous presence in several industries due to their availability and low price. Bus super-structures are typically built using structural steel hollow shapes and serve a major role during crashes and rollovers, as they protect the passengers by absorbing the kinetic energy of impacts and dissipating it as plastic deformations. In recent years, composite materials have gained protagonism in numerous applications due to their high specific strength and stiffness. However, costs and manufacturing complexity have made all-composite automotive structures economically unfeasible. Thus, the current tendency is the use of multimaterial structures: using composites only in the zones where they are needed, while keeping an inexpensive material, like steel, elsewhere. Hollow structural shapes, used in bus structures, are susceptible to bending collapse failure during rollover and crashes, which must be precisely predicted and calculated. Existing theoretical models for this failure mechanism have certain limitations to account for larger thickness, plastic hardening, and composite reinforcements. The present work aims to address these limitations through the development of new theoretical models for the so-called medium-thin-walled hollow shapes, as well as for reinforced CFRP-Steel hollow shapes. Both materials are joined using structural adhesives due to their ease-of-use and relatively low price. Experimental test results have shown the validity and accuracy of the proposed models. These proposed models are then implemented in a concept model of a bus structure to address its crashworthiness and the effectiveness of the reinforced shapes.