Parents of children with cerebral palsy (CP) have desires for changes in their children's mobility. Knowing the level of mobility of these children is crucial for guiding the prognosis and aligning family expectations.

To identify the development of mobility performance in Brazilian children and adolescents stratified by each level of the Gross Motor Function Classification System (GMFCS).

This is a multicenter study, which included 313 Brazilian children and adolescents with CP (aged 2 to 18 years). Mobility performance was assessed by the Gross Motor Function Family Report (GM-FR), and the classification level was determined by the GMFCS. To identify mobility performance, the nonlinear regression model developed by Palisano (2000) was used.

A total of 368 evaluations were performed, with a mean age of 7.21 years (±3.746). The GMFCS level proved to be a strong predictor and explained 78% (R² = 0.78; p < 0.001) of the variation in mobility, while age was not a predictor (R² = 0.23; p < 0.061). The predicted GM-FR scores differed across the five mobility curves: [level I (N = 63) = 89.1; level II (N = 70) = 80.6; level III (N = 40) = 56.7; level IV (N = 60) = 32.8; level V (N = 128) = 13.9]. The difference in the predicted GM-FR score between the GMFCS levels was, on average, 19.88 points, with the smallest difference between levels I and II (8.9 points). The age at which the children reached 90% of the predicted GM-FR score was: GMFCS I: 4 years; GMFCS II: 6 years and 1 month; GMFCS III: 3 years and 6 months; GMFCS IV: 3 years and 9 months; GMFCS V: 3 years and 11 months.

The classification of children with CP based on abilities and functional limitations is predictive of mobility. Children with GMFCS levels I and II exhibit less variation in mobility between them, while the other levels show greater variation. Additionally, children with higher GMFCS levels take longer to reach the expected limit.

The implications of this study highlight that the functional classification of children with cerebral palsy, based on the GMFCS, is a strong predictor of motor performance and mobility. Children in GMFCS levels IV and V reach the expected mobility limit more quickly, in contrast to children with GMFCS levels I and II, who show less variation in performance. This suggests that, for children with higher GMFCS levels, therapeutic interventions should focus on maximizing functional capacity within their level of mobility, given that they face greater limitations. Furthermore, everyday performance skills, as reflected in the GM-FR, emerge as stronger predictors of mobility than chronological age. This finding underscores the need for rehabilitation strategies that consider functionality in real-life contexts, not just motor skills in clinical settings, and highlights the importance of public policies that promote inclusion and accessibility for all children with CP, especially those with GMFCS levels IV and V.