The growth curves of the children are the most commonly used tools to assess the general welfare of society. Particularity child being one of the pillars to develop society; through these tools, we can path a child's growth physiology. The Centile line is of the important tools to build these curves, which give an accurate interpretation of the information society, also respond with illustration variable age. To build standard growth curves for BMI, we use BMI as an index. LMSP method used for finding the Centile line which depends on four curves represents Median, Coefficient of Variation, Skews, and Kurtosis. These can be obtained by modeling four parameters as nonparametric Smoothing functions for the illustration variable. Maximizing penalized likelihood in the Fisher scoring algorithm includes RS& GS algorithms. The resulting centiles expressed multi-level of body mass index ranging from lack of nutrition to obsess, also LMSP gives good results, and BMI with positive skew increases with age. Data on the variable applied to Iraqi children under five years of the multiple indicator cluster survey in 2011 and under the supervision of the world health organization (WHO).
