This study examined whether children with hemiplegic cerebral palsy (CP) have anticipatory control of fingertip forces during lifts of familiar objects, and what type of practice (blocked or random) best enhances the retention of anticipatory control during lifts of novel objects. Eighteen children with hemiplegic CP (7 females, 11 males; 7 to 14 years of age, mean age 10 years, SD 1.8) and 18 age-matched typically developing children (8 males, 10 females; mean age 10.4 years, SD 1.7) participated in the study. In the first experiment the children lifted familiar objects of various weights and sizes five times each, while the vertical lifting (load) force was measured. Most participants demonstrated higher rates of load force increase for heavier (and larger) objects already during the first lift, indicating anticipatory control. Furthermore, the load force rates generally were similar across the five lifts for each object, suggesting that they had stable representations of the objects' properties. In the second experiment children lifted three novel objects varying in weight (but identical in volume) 27 times each, in either a blocked or a random order, followed by nine immediate and nine delayed (24 hours) retention trials. Blocked practice resulted in greater differentiation of the force rates between objects during acquisition than did random practice. Both practice schedules resulted in similar retention. These findings suggest that children with hemiplegic CP have a priori internal representations used for anticipatory force scaling with familiar objects. Furthermore, the results indicate that these children can form and retain internal representations of novel objects for anticipatory control, irrespective of the type of practice schedule employed. Thus, clinically based practice sessions that incorporate lifts with novel objects may enhance anticipatory force scaling and related prehensile function in children with hemiplegic CP.