During the early stages of embryonic development, the heart is a smooth-walled, muscle-wrapped tube that bends and rotates in a vital, but poorly understood, morphogenetic process called looping. Since looping involves biomechanical forces, this paper examines two mechanically based hypotheses for the bending component of cardiac looping. The first hypothesis is that an initial tension in or near the dorsal mesocardium (DM), a longitudinal structure along the outside of the ventricle, drives the deformation. To relieve the bending stresses in the tube, the myocytes change shape passively, and then they deform actively to continue the process to completion of a full loop. In the second hypothesis, contraction of circumferentially arranged actin macrofilaments produces circumferential compression and longitudinal expansion (due to incompressibility) of the myocytes. The DM locally constrains the longitudinal deformation, forcing the tube to bend. The feasibility of these hypotheses was evaluated using theoretical models and published experimental results. The models, which consist of beams composed of two layers representing the DM and the ventricular myocardium, show that the hypotheses are consistent with most of the known data, but further studies are necessary. In this regard, the models provide a conceptual framework for designing experiments to investigate the mechanics of looping.