We trace the history of attempts to determine whether the experimentally observed diminution of metabolic energy expenditure when muscles lengthen during active contraction is consistent with reversibility of biochemical reactions and, in particular, with the regeneration of ATP. We note that this scientific endeavor has something of a parallel flavor to it, with both early and more recent experiments exploiting both isolated muscle preparations and exercising human subjects. In tracing this history from the late 19th century to the present, it becomes clear that energy can be (at least transiently) stored in a muscle undergoing an eccentric contraction but that this is unlikely to be due to the regeneration of ATP. A recently developed, thermodynamically constrained model of the cross-bridge cycle provides additional insight into this conclusion. Copyright © 2010 the American Physiological Society.