The effect of spastic cerebral palsy on in vivo gastrocnemius muscle fascicle length is not clear. Similarity of fascicle lengths in children with diplegia and typically developing children, but shortening of fascicle lengths in the paretic legs of children with hemiplegia compared with the non-paretic legs, are both reported. In the former case, comparisons were made between fascicle lengths normalized to leg length, whereas in the latter case, absolute fascicle lengths were compared. The inherent assumptions when normalizing fascicle length (measured via ultrasonography) were not validated, raising the possibility that inappropriate normalization contributed to the controversy. We used statistical methods to control the potential confounding effect of leg length on fascicle length, and tested the feasibility of the normalization method for a group of 18 children with diplegia (nine males, nine females; mean age 8y 7mo [SD 3y 11mo], range 2-15y; Gross Motor Function Classification System levels II and III) and 50 typically developing children (20 males, 30 females; mean age 9y 1mo [SD 2y 4mo], range 4-14y). Children with diplegia, as a group, had shorter absolute and normalized fascicle lengths (p<0.05) but we could not refute the appropriateness of the normalization method. Other methodological issues (such as sample characteristics) might have contributed to the apparent controversy between the studies.