The present study examines the effects of a four-week training session in a pegboard task on uni- and bimanual performance. Of particular interest was whether practice transfer from the trained to the untrained hand takes place. Twenty-five consistently right-handed subjects were trained to perform the task with the dominant hand, or the subdominant hand, or with both hands. After this training, the learning effects for the trained and untrained hands were analysed. To summarise, we obtained the following findings: (1) After training, movement times were considerably reduced for all hands and for all training conditions (practice effects); (2) practice effects were found for the hand trained and also for the untrained hand; (3) there was not a great difference in the size of the practice effects for the right hand after left hand training or for the left hand after right hand training; (4) task difficulty had no clear influence on the practice effect; (5) and finally, we discovered that bimanual movements not only profit from bimanual training but also from unimanual training and conversely unimanual movements benefit from bimanual training. These findings are discussed in the context of different motor control models and in the light of recent brain imaging findings.