Abstract The aim of this paper was to test the claim of the Visual Attention Model (VAM, Schneider, 1995) that an object that is visually attended to for the purpose of perceptual processing is the only object selected-for-action. Experiment 1 aimed to test this claim directly whilst Experiment 2 extended it to see if selection-for action is also bound to the target of stimulus-driven attention. A dual-task paradigm was used in which participants were required to discriminate (Experiment 1) or detect (Experiment 2) a target to determine whether a goal-directed pointing movement should be made. On a minority of trials the movement target could jump at movement onset to either the left or the right. The efficiency of the automatic pilot at making corrections to this jump was used as an operational measure of selection-for-action. Contradictory to the VAM the location of attention, whether it was oriented endogenously or exogenously, was found to have no effect on these operational measures. The results of the paper indicate a remarkable independence of selection-for-action from both selection-for-perception and stimulus-driven attention.