Aim: To examine the passive length–tension relations in the myotendinous components of the plantarflexor muscles of children with and without cerebral palsy (CP) under conditions excluding reflex muscle contraction. Method: A cross-sectional, non-interventional study was conducted in a hospital outpatient clinic. Passive torque–angle characteristics of the ankle were quantified from full plantarflexion to full available dorsiflexion in 26 independently ambulant children with CP (11 females, 15 males; mean age: 6y 11mo, range 4y 7mo–9y 7mo) and 26 age-matched typically developing children (18 females, 8 males; mean age 7y 2mo, range 4y 1mo–10y 4mo). In the children with CP, the affected (hemiplegia; n=21) or more affected (diplegia; n=5) leg was tested; in typically developing children, the leg tested was randomly selected. Gross Motor Function Classification System levels were I (n=15) and II (n=11). Care was taken to eliminate active or reflex muscle contribution to the movement, confirmed by the absence of electromyographic activity. Results: There were small but significant differences between the two groups for maximum ankle dorsiflexion (p=0.003), but large and significant differences in the torques required to produce the same displacement (p<0.001). Further, the hysteresis of the average loading cycle in the children with CP was over three times that of the typically developing children (p<0.001). Interpretation: We believe that the plantarflexor muscles of children with CP are stiffer and intrinsically more resistant to stretch, even though they retain near normal excursion. This increased stiffness is a non-neurally-mediated feature demonstrated by these children. The extent to which it influences function and predisposes the children to development of soft tissue contracture is unknown.