The purpose of this experiment was to examine the effects of advance knowledge on the kinematics of one-handed catching. Balls were launched from a distance of 8.4 m by a ball-projection machine with adjustable launching speed. Fifteen skilled ball catchers caught 160 balls with their preferred hand under blocked-order (4 blocks, each comprising 20 trials at 1 of 4 different ball speeds) or random-order (4 blocks, each comprising 20 trials of 4 different ball speeds) conditions. By projecting balls with different ball speeds from a fixed position, it was possible to modify the temporal constraints of the catching task. In both the blocked-order and random-order conditions, catching performance (number of catches, touches and misses) decreased with increasing temporal constraints. Analysis of successful trials indicated that this equal level of catching performance was achieved with different movement kinematics. Specifically, there was a change in movement time, latency, wrist velocity profile, and coefficient of straightness. Based on expectancy of previous trials, movement kinematics was scaled to ball speed in the blocked-order condition whereas in the random-order condition, participants exhibited a more default initial response. However, this latter mode of control was functional in that it increased the likelihood of success for the higher ball speeds while also providing participants with a larger temporal window to negotiate the unexpected temporal constraint on-line for the lowest ball speed. © 2009 Springer-Verlag.