A movable platform designed to study balance (perturbation paradigm) was used to investigate typically developing (TD) children and those with cerebral palsy (CP). Previous research typically used small and predictable threats to balance (Burtner, Woollacott, & Qualls, 1999); because of small perturbation sizes behavior strategies were somewhat similar among TD and children with CP (Burtner, Woollacott, & Qualls, 1999; Burtner, Qualls & Woollacott, 1998). As a result little is known about response strategies demonstrated by children with CP. How do children with CP manage the forces to regain balance in changing conditions? Thirty-nine TD children (n=39), 9 months to 10 years, and eight (n=8) who have CP (spastic diplegia) between 2 to 10 years, were studied to delineate differences in reactive responses to increasing perturbation magnitudes. Backward support-surface translations (tri-axial force plate embedded in moveable platform) graded as easy, moderate and difficult were imposed and comparisons were made between developmentally- and chronologically-matched groups. Developmental groupings were based on performance in balance and locomotor tasks: a) Pre-walkers (PWK; CP: n=3; TD: n=2), b) New-Walkers (NWK; CP: n=3; TD: n=6) and c) Runners/jumpers. (RNJ; CP: n=2, TD n=8). Age-matched groups consisted of children 5-years and under (CP: n=3; TD n=9) and children 6-10 years (CP: n=5; TD n=16). Significant differences were found between groups in both amplitude and velocity thresholds: perturbations where falls resulted or steps and assistance were needed. Those with CP had lower percentages of feet-in-place responses but greater percentages of step- and loss-of-balance responses compared to TD as task difficulty increased. Children with CP also showed greater center-of-pressure (COP) excursion relative to optimum COP trajectory (COPPath), increased frequency of directional COP changes, and longer recovery time (COP stability). Overall, children with CP showed a reduced response capacity with increasing balance threats. Both chronological and developmental comparisons showed significant differences between groups for most variables examined suggesting not just developmental delays, but other factors such as spasticity and musculoskeletal configuration as contributors to underlying balance performance. By challenging children with CP to the limits of their abilities (vs. small perturbations in previous research) this study clearly distinguished TD children from those with CP. In both clinical and laboratory settings, challenging children with CP, will determine their motor capacity and extent of impairment in reactive balance control. Determining the threshold (velocity or amplitude) at which a child takes a step also helps monitor improvement during therapy (pre- vs. post-balance training). Keyword(s): assessment, early childhood, research