The goal was to compare children with hemiplegia with those with diplegia within Gross Motor Functional Classification System (GMFCS) levels using multiple validated outcome tools. Specifically, we proposed that children with hemiplegia would have better gait and gross motor function within levels while upper extremity function would be poorer. Data were collected on 422 ambulatory children with cerebral palsy: 261 with diplegia and 161 with hemiplegia, across seven centers. Those with hemiplegia in each level performed significantly and consistently better on gait or lower extremity function and poorer on upper extremity and school function than those with diplegia. In GMFCS Level II, the group with hemiplegia walked faster (p=0.017), scored 6.6 points higher on Dimension E of the Gross Motor Function Measure (p=0.017), 6.7 points lower on Upper Extremity subscale of the Pediatric Outcomes Data Collection Instrument, and 9.1 points lower on WeeFIM self-care (p=0.002). Basing motor prognosis on GMFCS level alone may underestimate lower extremity skills of children with hemiplegia, and overestimate those of children with diplegia.