Naive human subjects (N=18) were asked to move a manipulandum on a plane in directions other than going straight towards a visual stimulus. They were instructed verbally to generate a movement at an angle from a stimulus direction which varied in 2-dimensional (2-D) space from trial to trial in a pseudorandom fashion. Each subject performed eight sets of twenty consecutive trials: one for moving in the stimulus direction and seven for moving in directions at an angle from it. The angles were 5, 10, 15, 35, 70, 105 and 140°. Nine subjects were instructed to move in the clockwise (CW) departure and 9 to move in either (EI) the clockwise or the counterclockwise (CCW) departure, as they wished. The direction of the movement in 2-D space and the reaction time (RT) were measured. The mean angle achieved in a given set overshot the instruction angle, especially in the lower range (5-35°). The reaction time, (RT₀), of movements made at an angle from the stimulus direction showed two kinds of change: first, a step increase from the reaction time, RT0, of movements in the stimulus direction, and second, superimposed upon it, a linear increase with the amplitude of the angle. The slope of the line was similar for the CW (2.37 ms/degree) and the EI case (2.28 ms/degree), but the step increase (y-intercept) for the EI case (84 ms) was substantially less than that of the CW case (155 ms). The linear increase of the RT with angle is compatible with the idea that performance in the task may involve a mental rotation of the imagined movement vector about its origin. The rotation would begin from the stimulus direction and end when the required angle is judged to have been reached; in addition, corrections of this angle at the end of the rotation could be made. The slope of 2.37 ms/degree observed in the CW case would correspond to a rotation rate of 422 degrees/s. The finding of a similar rate for the EI case indicates a similarity in strategy with regard to achieving a desired angle. In contrast, the lower intercept observed for the EI case suggests significant savings in processing information which is unconstrained with regard to angular departure. Assuming this model of internal motion, we analyzed the amplitude-accuracy relations using Fitts' (1954) approach to real movements. In accordance with Fitts' law, we found that the increase in RT, considered as a mental movement time, was a linear function of task difficulty which was calculated from the angle achieved and its variability. This indicates that Fitts' law holds for the hypothesized rotatory motion of the imagined movement vector, and that both real and imagined movements might be governed by similar amplitude-accuracy relations.