Background: The position of the gastrocnemius tendon in relation to the leg length may be different in children with cerebral palsy as compared to normal children. The palpation of muscle bellies or previous experience of the operating surgeon is employed to place the surgical incision for lengthening of the gastrocnemius aponeurosis. Inaccurate localisation may cause incorrect incisions and a risk of iatrogenic damage to the vital structures (i.e. sural nerve). Objectives: The aim of our study is to compare gastrocnemius lengthin-vivobetween paretic and unaffected children and create a formula to localise the muscle-tendon junction accurately. Methods: 10 children with di/hemiplegia (range 2-14y) were recruited. None of them had received any conventional medical treatment. An equal number of age/sex matched, typically developing children (range 4-14y) were recruited. Ultrasound scanning of the gastrocnemius muscle at rest was performed to measure the length of gastrocnemius bellies. We also measured the heights and leg lengths in all the children. Results: The gastrocnemius medial muscles were shorter in Cerebral Palsy children when compared to similar aged normal children. In cerebral palsy children, the gastrocnemius muscle and leg ratio ranged between 35 to 50% (average ratio of 45%). Conclusion: Using these figures, we created an average percentage for gastrocnemius muscle length that may be used clinically to identify the tendon for open/endoscopic lengthening and also to make simple and accurate localisation of gastrocnemius muscle-tendon junction for surgical access. This decreases the length of the surgical incision and may reduce the risk of iatrogenic injuries.