Echocardiographic measures of mitral annular descent (MAD) assume a fixed left ventricular (LV) apex throughout the cardiac cycle, ignoring the apical component of LV long-axis shortening (LAS). We tested whether apical motion contributes significantly to LAS, making LAS a better surrogate of LV systolic function than MAD. Three-dimensional LV systolic MAD, LAS, and apical motion were measured in sheep using implanted radiopaque markers and biplane videofluoroscopy. End-diastolic volume-stroke work relationship (preload recruitable stroke work) was computed as a load-independent index of LV systolic function. Apical motion was 1.4 +/- 0.8 mm, representing 22% of LAS (P <.05). Linear regression demonstrated that MAD correlated slightly better with preload recruitable stroke work (r = 0.808) than LAS (r = 0.792, both P <.001). Receiver operating characteristic curves demonstrated MAD was more accurate in predicting depressed LV function than LAS (93% vs 84%, respectively). Although LV apical motion contributed significantly to LAS, MAD measured with a fixed-apex assumption, as currently done echocardiographically, correlated more closely with LV preload recruitable stroke work.