Quantitative ultrasound of muscle echogenicity may be influenced by the size of the muscle and, so, corrections may be required when comparing echogenicity between populations with different muscle size. This study examined the relationship between regional muscle echogenicity and geometry in young, typically developing children (n = 49, mean ± standard deviation [SD] age = 70.8 ± 30.0 mo). Three-dimensional ultrasound was used to measure mean echo intensity (EI) and echo variation (EV), together with muscle thickness, cross-sectional area (CSA) and width, over the entire muscle length for the medial gastrocnemius (MG), rectus femoris (RF) and tibialis anterior muscles. Pearson's correlation coefficient (r) was used to assess the strength of the relationship between echogenicity and geometry using all images taken over the entire muscle length. There were moderate-strong correlations (r = 0.67–0.90) between EI and EV and thickness for each muscle, with the strongest correlations evident for the MG and RF. EI and EV were moderately correlated with muscle CSA and weakly correlated with muscle width. Normalisation of echogenicity to muscle thickness may help delineate between regions of contractile and non-contractile tissue and provide a useful measure of muscle echogenicity when comparing muscles of different cross-sectional dimensions. We recommend that researchers consider controlling for muscle size-dependent effects on echogenicity, by normalising EI and EV to muscle thickness, or including the latter as a covariate when comparing between groups.