Measurements of segmental contraction of the left ventricle by standard magnetic resonance imaging (MRI) and two-dimensional echocardiography involve the comparison of diastolic and systolic time frames acquired from the same imaging plane in space. As the cone-shaped left ventricle shortens along its long axis during systole, the observed contraction may differ from the true myocardial contraction. Spin-echo MRI examinations in 21 healthy subjects were performed to evaluate the error caused by failing to compensate for through-plane motion. The authors found that at the base and the mid-ventricle the observed contraction systematically underestimates true contraction by an average of 16% and 21%, respectively (P less than .001). At the apex, the segmental contraction may be overestimated or underestimated. Because of this error, standard MRI and echocardiography are less suited for basic research on cardiac contraction patterns. However, standard imaging techniques are valuable in clinical studies comparing groups of patients, because all measurements will suffer from the same systematic error.