The present mirror-drawing experiment was conducted over four months on one healthy female (R.H.) to elucidate the nature of the ipsilateral and contralateral learning curves for 12 sessions of unimanual training of the nonpreferred hand using a random practice schedule. Analysis indicated greater initial improvement for the practising nonpreferred hand with further improvements for both hands in number of errors and time to complete a single transfer test trial. A 4-mo, retention test showed the improvements were greater for the preferred (untrained) hand. Star-tracing records completed for both hands were quite similar to those recorded using the mirror after practice. This may reflect central processing and long-term potentiation influential even after practice ceased.