The present study investigated the mechanisms involved in the preparation of pointing movements in humans. We provided visual precues on the location of the upcoming target, and registered the effect of these precues on the reaction time (RT = interval between target appearance and movement onset). Generally, precues were found to reduce RT, suggesting that some aspects of the preparatory process have been advanced in time. In Exp. 1, precues fully specified the direction required for the upcoming movement while indicating only a range of movement amplitudes; in Exp. 2, precues fully specified the amplitude and indicated a range of directions. In both experiments, RT was shorter than in control trials without precues, and gradually increased with the size of the precued amplitude or direction range. This result suggests that the preparation of either parameter is possible without knowing the precise value of the other, i.e. amplitude and direction are not prepared in a fixed order. Furthermore, our results are consistent with the view that movement preparation includes a progressive contraction of the precued range towards the final value. The speed of this process can be estimated as 0.31 cm/ms for amplitude, and 1.7 deg/ms for direction ranges. In Exp. 3 and 4, precues indicated both amplitude and direction as ranges only. The size of the amplitude range was held constant while the size of the direction range was varied (Exp. 3), or vice versa (Exp. 4). Under these conditions, RT increased with the size of the varied range.(ABSTRACT TRUNCATED AT 250 WORDS)