Maternal depression has been associated with another cumulative risk conditions with negative impact for school-age children outcomes highlighting the mediate effect of poverty and single parents for that association. However, few studies included in the explicative models these variavble together as a cumulative risk. The aim of this study was identify the weight of sociodemographic variable for the behavior of school-age children who live with maternal recurrent depression and explore explanatory models that include cumulative sociodemographic varibles. In a cross-sectional, predictive study, it was included 100 mothers and school-age children (7-12 years), 50 mothers with recurrent depression (clinical sample) and 50 motehrs with no mental disorders (community sample), which were systematically evaluated by a diagnostic interview (SCID) and answered the SDQ to identify children behavioral problems. Mothers answered yet a sociodemographic questionnaire and two stressful events scales, Brazilian instruments with satisfactory psychometric properties. We proceed to satatiscal analysis, including linear regression, in three models: Model 1 - with maternal depression as independent variable, Model 2 - included maternal education and occupation, number of children, family income and composition and socioeconomic class, Model 3 - included also scores of stressful events scales. Maternal depression was associated in this sample, with statistically meaningfulness, to almost of the tested variables, with moderate correlations, which allowed the inclusion of these variables in the other models. The estimated regression models were signifcant, in the bivariate linear regression analysis with maternal depression [F(1,99)=912.040, p<.001, R2a=.160)] and in the multiple linear regression for Models 3 and 2, where maternal education was the variable with significant weight. The best model tested was the third one, that includes maternal depression, sociodemorgaphic variables and cumulative stressful events [F(9,99)=1687.656, p<.001, R2a=.223]. These findings highlight the relevance of identify cumulative risks, including low maternal education, as guiding practices in children´s mental health.