An experiment tested the hypothesis that fast, short sequences of movements are planned as a whole, before movement inception. The experimental task consisted of pointing to either one (one-step condition), or two (two-step condition) visual targets aligned along the midsagittal axis in a horizontal plane. There were nine possible arrangements of the targets resulting from all combinations of three distances (5, 10, 15 cm), and two trial orders (blocked or random). Performances were characterised by reaction time (RT), movement kinematics, and spatial accuracy. Compared with one-step trials, the first movements of two-step trials had longer RTs (length effect), particularly in random sessions, and when the sequences included short-distance targets. There were also differences in duration (one-target advantage), velocity profile and spatial accuracy that did not depend on the characteristics of the second movement. The results are inconsistent with the assumption that two-step sequences are planned as a whole. Instead, they are in keeping with the alternative hypothesis that part of the preparation of the second step takes place during the execution of the first step.