A nutritional status index was built by modeling the mathematical function of the mean Z scores of weight for age, from 60,079 children under five years of age, selected in a probabilistic fashion from the Mexican population. The most precise mathematical model was a fifth degree polynomial. The correlation coefficient was between .937<R<.999. The proposed index is the integral of the polynomial function, which represents the nutritional gap between the observed and the reference population. This model is used to analyze the characteristics of the different study populations. The index shows an improvement of -39.6 to -16.8 at the national level, between the years 1988 and 1999. This improvement is greater in urban (-36-4 to -8.4) than in rural areas (-48.8 in 1989, to -37.7 in 1999). The indigenous rural population of the country showed the highest levels of malnutrition (-54.4), as compared to the non-indigenous rural population (-37.1). In Mexico, City, the index was -5.9 in 1995, which represents an average of extreme nutritional values: -17.3 in the lower socioeconomic strata and 18.0 in the higher strata, the latter suggesting the presence of childhood obesity. This index is useful to estimate the epidemiological burden and the characteristics of malnutrition at early ages, as well as to assess the impact of interventions, without being altered by common biases related to the utilization of malnutrition prevalence values.