Although most cardiac cell therapy trials have focused on patients with acute myocardial infarction, attempts at "regenerating" chronically failing hearts have also been performed. These studies have entailed use of skeletal myoblasts and bone marrow-derived cells. In the case of skeletal myoblasts, the randomized placebo-controlled MAGIC trial has not achieved its primary end point as 6-month ejection fractions did not significantly differ between patients receiving cells or placebo, but the finding that the highest dose of myoblasts resulted in a significant anti-remodeling effect (a prespecified secondary end point) compared with the placebo group provides an encouraging signal. In the case of bone marrow cells, surgical injections of the mononuclear fraction combined with coronary artery bypass surgery have failed to show any substantial benefit. A catheter-based trial using a cross-over type of design has reported more successful outcomes, but its results will then have to be confirmed. Indeed, the most positive results have been reported with intraoperative epicardial injections of CD133 progenitors, which is probably explained by the angiogenic potential of these cells. There are three possible reasons for these mixed results. The first is the marked heterogeneity of cell functionality (particularly in the case of bone marrow), which would expectedly translate into variable clinical outcomes. The second reason is the low rate of sustained engraftment caused by early mechanical leakage followed by biologically induced cell death. The third possible explanation is a mismatch between the choice of end points and the presumed mechanism of action of the cells. The initial assumption that adult stem cells could affect myocardial tissue regeneration has led to the usual focus on ejection fraction as the major surrogate end point for treatment efficacy. It is now increasingly recognized that adult stem cells, in contrast to their embryonic counterparts, have little if any regenerative capacity and that their presumed beneficial effects more likely involve paracrine signaling and/or limitation of remodeling, in which case infarct size, perfusion, or left ventricular volumes might be more appropriate markers. Altogether, these observations provide a framework for future research the results of which will then have to be integrated into the protocol design of second-generation clinical trials so as to maximize their likelihood of yielding more successful results.