The aim of the study was to determine the possible correlation between the degree of femoral anteversion and the quantitative data obtained by 3D Gait Analysis (GA) and then to investigate the relationship between femoral anteversion and the reduced knee flexion during swing phase in children with Cerebral Palsy. Twenty-seven diplegic children with severe rectus femoris spasticity and 20 healthy children (CG) were considered. Clinical evaluation of femoral anteversion, Duncan Ely test and Gait Analysis were performed in all patients. From Gait Analysis data some indices were identified and calculated and statistical analysis performed. Clinical evaluations made the distinction between patients with excessive femoral anteversion (Group 1) and those with normal value (Group 2). Both groups showed a blunt maximum of knee flexion in swing (KMSw), representative of rectus femoris spasticity, but two different gait strategies were found for the timing of KMSw. Group 1 exhibited a reduced KMSw value with its timing close to normal value and an excessive hip internal rotation (Mean Hip Rotation index), correlated to high femoral anteversion; Group 2 presented a limited KMSw and a significant delay of its timing, with Mean Hip Rotation index close to Control Group. No differences were found for other indices. The results demonstrated that the presence of reduced KMSw only can be directly connected to excessive femoral anteversion; the coexistence of reduced KMSw and its delayed timing reveals that the rectus femoris spasticity may be due to rectus spasticity added to an incorrect motor selective control. The results are clinically crucial for treatment strategies (derotative femoral osteotomy vs rectus transfer).