BACKGROUND: As the benefits of extemporaneous transplantation (Tx) of fresh (unfractionated) autologous bone marrow (BM) have been primarily studied in the setting of acute myocardial infarction, we assessed whether this approach could be effective for regenerating chronically infarcted myocardium. METHODS AND RESULTS: Myocardial infarction was created in 18 sheep by ligation of circumflex arterial branches. Three weeks later, BM was aspirated from the iliac crest, washed, labeled with the fluorescent dye Dil and reinjected (mean: 422 x 10(6) cells in 3 mL) in 10 sites across the infarcted area through the reopened thoracotomy (n=9). Nine controls received culture medium. Left ventricular (LV) function was assessed before and 2 months after Tx by two-dimensional echocardiography whereas transmural velocity gradients were measured using M-mode tissue Doppler imaging at the center of the infarcted/grafted area. Formalin-fixed hearts were processed for the detection of grafted cells and angiogenesis. LV ejection fraction deteriorated similarly in the Tx and control groups (from 42+/-5% to 30+/-4% and from 40+/-4% to 31+/-1%, respectively; P=0.86). Likewise, BM Tx failed to prevent LV dilatation and impairment of the global wall motion score. The decrease in regional systolic velocity gradients (s(-1)) featured a similar pattern (Tx group: from 0.77+/-0.11 to 0.31+/-0.07; control group: from 0.73+/-0.10 to 0.50+/-0.07; P=0.06). Histologically, there was neither BM tissue engraftment, except for a few scattered Dil-positive macrophages in the infarcted fibrotic areas nor transdifferentiation of BM cells into endothelial cells. CONCLUSIONS: These data caution against the functional efficacy of extemporaneous Tx of fresh unfractionated BM into postinfarction scars.