The role of movement amplitude and target diameter in motor programming and reaction time in a target tap task in Fitts’ paradigm is not clear. Fitts and Peterson (1964) reported that reaction time increased as movement amplitude increased. In contrast, Klapp (1975) reported that reaction time increased with decrease in target diameter. Alteration in limb segment recruitment with alteration in movement amplitude, and pooling reaction time observations from responses to different directions and amplitudes in the same analysis might have contributed to the apparent conflict in the pervious work.

We report an experiment, employing a discrete Fitts’ task, in which twelve participants used a long stylus that allowed them to use similar sets of limb segments to tap targets of different diameters that were located at different amplitudes (short or long). Another group of twelve participants used a conventional stylus. Therefore, limb segments were free to vary according to the task constraints or movement strategies. In addition to the analysis of pooled data, reaction time and premotor time were separately analysed for movements to the left and right, and for the short and long amplitudes.

For the short-stylus group, the results confirmed those reported by Klapp (1975). In the long-stylus group, reaction time and premotor time increased with decrease in target diameter. Reaction time, but not premotor time, was affected by movement amplitude and increased as the amplitude decreased. The movement amplitude effect on response latency was influenced by the end-effector characteristics.

It is concluded that movement amplitude should be defined in terms of the monotonic rotation in the jointspace. Reaction time results provide an explanation for the apparent conflict between Fitts and Peterson’s and Klapp’s studies. Moreover, they are consistent with an interpretation of the Knowledge Model and Minimum Cost Hypothesis for the control of a manipulator with redundant degrees of freedom. Electromyographic findings revealed the use of speed-sensitive and speed-insensitive strategies by the participants in the long-stylus group.