Objective: To compare the effects of physical rehabilitation with (PRT) and without (CON) progressive resistance training following treatment of spastic plantarflexors with botulinum toxin type A (BoNT) in children with cerebral palsy (CP). Methods: Fourteen children with CP performed supervised PRT (n=7) or CON (n=7) two times per week for 12 weeks, following the BoNT-treatment. Outcome measurements were performed at baseline (pre BoNT), and 4 and 12 weeks post BoNT. They consisted of: ankle muscle function (maximal torque and submaximal torque steadiness of isometric ankle dorsi- and plantarflexion and associated ankle muscle [EMG] activity), gait function (3-dimensional gait analysis), balance function (sway analysis), gross motor function (GMFM-66), and spasticity (modified Ashworth). Results: Submaximal torque control (torque steadiness) of isometric dorsiflexion improved similarly in the two groups, and the improvement was related to the reduction in antagonist (soleus) co-activity (P<0.05). Maximal plantarflexion torque increased after PRT, whereas a reduction was seen after CON (P<0.05). No changes in function were observed. Conclusions: Both types of physical rehabilitation in combination with BoNT-treatment improved antagonist (ankle dorsiflexion) torque-control to the same extent - which was related to the reduction in antagonist co-activity - but only rehabilitation with PRT increased maximal plantarflexion torque.