The "Bolivian orocline" is the change in trend of the Andes from NW to N near 18°S. Paleomagnetic data have been used to infer that this bend was in part produced by wholesale opposed rotations (15-20° counterclockwise north of the bend and 15-20° clockwise south of it) of the two limbs of the orocline. Besides paleomagnetic data, shortening estimates from balanced cross-sections and variations in crustal cross-sectional area provide quantitative information on rotations and translations. The three sets of data do not agree closely, and therefore only loosely constrain the kinematics of the orocline. A map view kinematic model incorporating the major faults gives a more detailed picture. The resulting displacement field suggests that movement perpendicular to the orogen increases toward the bend, whereas the component of orogen-parallel movement increases away from it. Only weak rotations are indicated for the bend region. It is speculated that large-scale regional rotations of the limbs of the orocline are markedly lower than previously suggested, probably of the order of 5-10°. Stronger rotations are associated with strong lateral shortening gradients or may result from superimposed local phenomena.