Left ventricular fractional shortening (FS) is dependent on left ventricular preload and afterload, as well as contractility. Contractility may therefore not be accurately described by FS, especially in infants and children infected with human immunodeficiency virus (HIV), who tend to have abnormal left ventricular preload and afterload. We therefore examined the magnitude and clinical impact of the discrepancy between FS and contractility by assessment of 177 echocardiograms from 76 HIV-infected pediatric patients (median age, 1.9 years). The studies included simultaneous measurements of left ventricular FS, contractility, preload, and afterload. The correlation between contractility and FS was modest (r = 0.70; p < 0.001), and was weaker in children less than 2 years of age (r = 0.52) than in older children (r = 0.84). FS incorrectly predicted contractility in 46% of the studies; 43% with depressed FS (< 28%) had either normal (17/42) or enhanced (1/42) contractility. For 67% of echocardiograms, FS and contractility differed by > 1 SD, and for 36% the difference was > 2 SD. These differences remained after adjustment of FS for age or body surface area. Afterload was abnormal in 42% and preload in 21% of all echocardiograms. High preload predicted that FS would overestimate contractility (p = 0.002); high afterload predicted that FS would underestimate contractility (p < 0.001). The discrepancy between FS and contractility was larger among children who were younger, had more advanced HIV disease, or were not sedated during echocardiography. One third of children with congestive symptoms had normal contractility and depressed FS; the discrepancy was primarily due to loading conditions. We conclude that the high incidence of abnormal loading conditions in HIV-infected infants and children limits the usefulness of load-dependent FS for assessing contractility. Measurements of loading conditions and load-independent indexes, which more directly reflect contractility, allow a more accurate determination of myocardial status and may lead to better clinical management.