AimThis study aimed to compare gross motor function between aquatic‐ and land‐based exercises in children and adolescents with cerebral palsy (CP).MethodsThe authors conducted an electronic search of nine databases from their inception to 21 November 2024 (PROSPERO registration: CRD42020194121). Inclusion criteria were randomized controlled trials involving aquatic‐based exercises for children and adolescents with CP, assessing gross motor function using standardized scales or tests. Three authors independently extracted data using a predetermined Excel form. The risk of bias was assessed with the PEDro scale. The body of evidence was synthesized using the GRADE approach. Meta‐analysis was conducted using the Revman 5.3 program.ResultsA total of 369 children aged 2–18 years from 15 studies were included. Most participants were ambulatory and classified as having spastic hemiparetic or diparetic CP. The majority of studies had a high risk of bias and small sample sizes. Aquatic‐based exercises were categorized as aquatic physical therapy, Halliwick, swimming exercises, gait training and exercises. Low‐quality evidence indicated that aquatic physical therapy resulted in higher gross motor function than land‐based exercises (SMD = 0.47, n = 93, 4 trials, I2 = 5%, p = 0.03), with a small effect size. No significant differences were found for Halliwick or swimming exercises.InterpretationThere is low‐quality evidence, because of high risk of bias, imprecision and inconsistency, suggesting that aquatic‐based exercises are comparable with land‐based exercises. Future research should focus on well‐designed interventions with adequate sample sizes to compare the combination of aquatic‐ and land‐based therapies with land‐based therapy alone.