Liepmann suggested that the left hemisphere contained the engrams for motor sequences. Other investigators have suggested that ideomotor apraxia may be caused by either a destruction of these engrams or a disconnection of these engrams from motor systems in the nondominant hemisphere. If these hypotheses are correct, then ideomotor apraxics should not only show a defect on previously learned motor tasks but also a defect in new motor learning. Nine right-handed, hemiparetic, aphasic apraxics were given six trials on a rotary pursuit meter. Eight right-handed, hemiparetic, aphasic, nonapraxic patients sewed as controls. All subjects were instructed to use their left (nonparetic) hand. The performance of the control group on the sixth trial was significantly better than that on the first trial, showing a distinct learning effect. In the apraxic group, however, there was no significant difference between the first and sixth trial, suggesting a defect in motor learning. This defect appeared to be caused by a combined defect of both acquisition and retention. © 1975 American Academy of Neurology.