This study explores the effect of women`s involvement in BRAC`s income and health development activities on the nutritional status of their children aged 6-72 months. MUAC measurements of 1,518 children aged 6-72 months (using TALC) were taken between April-August 1995 under the BRAC-ICDDR,B joint research project in Matlab using a four-cell-study design. Data analysis consisted of both bivariate and multivariate analysis, along with comparison with similar data from a baseline survey done in 1992. Findings reveal that prevalence of severe PEM has decreased significantly from 23.2% to 14.1% among children of BRAC member households (p<0.05) during the period between 1992 and 1995. However, among non-member households, the prevalence remained almost unchanged (21.2%). The positive effect of women`s involvement in BRAC development activities on their child`s nutritional status remained significant even after controlling for age and sex of the child; age, year of schooling and number of living children of the mother; per capita monthly expenditure; MCHFP area; and four study cells during regression procedure. The children, whose mothers were participating in BRAC development activities, were 41% less likely to suffer from severe malnutrition compared to those of non-member (p<0.05). However, gender differential in the prevalence of severe malnutrition was very pronounced among the children of BRAC member households (p<0.05). This may, in part, be explained by the fact that BRAG does not have gender-focused component in its programme. Programmatic implications of these findings are discussed.