Objectives: The study examined the association between an index of household wealth and key nutritional status indicators in children under age five in Armenia. The study compared the distribution of the child nutritional status indicators across socio-economic groups in 2000 and 2005. It also examined the correlation between the Wealth Index and subjective measures of socio-economic status (SES), as well as the relationship between all of those SES measures and child nutritional status in Armenia in 2005. Methods: Data were from the Armenia Demographic and Health Survey (DHS) conducted in 2000 and 2005. The analyses accounted for sampling features, including the stratification of the sample by regions and urban/rural areas and the primary sampling units, as well as the clustering of children within households. Three measures of undernutrition were examined: stunting (low height-for-age), wasting (low weight-for-age), and underweight (low weight-for-height). Analyses included chi-square, the Kappa statistic, unadjusted and adjusted logistic regression, and calculation of poor/rich odds ratios and concentration indices. Independent variables included the Wealth Index, an objective SES measure, and three subjective SES measures, respondents' perceptions about: "having enough money to meet needs," "making ends meet in the household," and "satisfaction with living space." Covariates included urban/rural residence, region, education in years for mothers and fathers, marital status of mothers, work status of mothers, mother's age in years at the time of the child's birth, mother's body mass index, child's age in months, child gender, birth weight in kilograms, and the number of months the child was breastfed. Results: The prevalence of stunting in the combined and weighted DHS populations for 2000 and 2005 was 17.4%, using the 2006 World Health Organization standards for child growth. The prevalence of wasting was 3.3%, and for being underweight 2.9%. In bivariate results, children in the second wealth quintile, those who were "poorer" but not the most poor, had lower rates of wasting and underweight than those in most of the richer quintiles. In adjusted analyses, none of the associations for the Wealth Index and child undernutrition indicators were statistically significant. Each additional year of a father's or partner's education was associated with significantly lower adjusted odds of stunting (OR 0.93, 95% CI 0.88-0.98) and underweight (OR 0.85, 95% CI 0.76-0.95). Each additional kilogram of the child's birth weight was associated with 53% lower odds of stunting (OR 0.47, 95% CI 0.35-0.63) and 72% lower odds of being underweight (OR 0.28, 95% CI 0.17-0.46). Armenian regions that had less favorable nutritional indicators were Gegharkunik (children had higher risk of stunting and underweight), Shirak (children had higher risk of wasting and underweight), and Vayots Dzor (children had higher risk of wasting and underweight). Residents in these three regions have poorer SES compared to those living in other Armenian regions. The calculation of poor/rich odds ratios showed a significant differential in the risk for stunting in 2000 (poor/rich OR 2.12; 95% CI 1.29-3.50), but no significant difference in 2005. The analysis of concentration curves and indices indicated a higher concentration of stunting and underweight in poorer households in 2000, and a slightly higher concentration in the richer quintiles in 2005. The "making ends meet in the household" indicator might be a better predictor of child undernutrition than the Wealth Index or the other two subjective indicators examined. Discussion: Findings suggest that the Wealth Index has limited ability to predict nutritional status of children in Armenia. Region and paternal education had highly significant associations with undernutrition; these results suggest that they are important independent socio-economic determinants of nutritional outcomes for Armenian children. The regional variation in malnutrition rates and malnutrition inequalities show the importance of examining community and regional level socio-economic variables in addition to individual and household level factors, and of targeting selected regions for further studies and public health interventions designed to improve child nutrition.