Twelve experienced male weight lifters of varying ability completed a series of bench press lifts at 95% of maximum. These lifts included a rebound bench press, which was performed without a delay between the downward and upward components of the lift, a bench press performed without a downward phase, and two bench press movements performed with various pause periods imposed between the downward and upward phases of the lift. Force and cinematographic data were collected during each lift. The augmentation to performance derived from prior stretch was observed to decay as a function of the pause duration. This relationship was accurately described (P less than 0.01) by a negative exponential equation with a half-life of 0.85 s. The nature of this decay is discussed with reference to the implications for stretch-shorten cycle movements that are performed with a period of pause between the eccentric and concentric phases and for stretch-shorten cycle research paradigms.