This study was undertaken using a case-control, prevalence survey design, aimed at verifying the effects of socioeconomic, environmental, and biological/morbidity variables on the growth of schoolchildren from low-income families. The study focused on schoolchildren whose height/age (H/A) ratio was < or = 2SD (NCHS). Control groups were paired according to sex into two groups: C1, whose H/A ratio was between -1SD (NCHS) and the median, and C2, with the H/A ratio > median to +1 SD (NCHS). The Mantel-Haenszel test was used to verify each variable with regard to the H/A ratio, while non-conditional multivariate analysis was used to identify which of the variable blocks had a significant effect and, in the following stage, to identify the variables with a significant effect within each block. In the first stage, the variables with a significant effect for C1 were socioeconomic. For C2, socioeconomic, environmental, and morbidity/biological variables were significant. In the final model, the remaining significant variables for C1 were socioeconomic, while those for C2 were socioeconomic and morbidity/biological. The authors conclude that socioeconomic variables are hierarchically superior to other risk factors.