To investigate the utility of moderate-to-severe anemia as an impact measure of malaria control interventions we analyzed data from countries that had two or more Demographic and Health Surveys or Malaria Indicator Surveys conducted between 2001 and 2011 containing information on insecticide-treated net (ITN) use and hemoglobin levels in children 6-23 months of age. Multivariate Oaxaca-Blinder decomposition for nonlinear response models with deviation contrast normalization for categorical variables were used to estimate the proportion of the decline in anemia prevalence over time due to increases in ITN use. Weighted average ITN use increased from 12.2% in baseline surveys to 44.3% in endline surveys and moderate-to-severe anemia decreased from 17.9% to 12.1%. In pooled, multi-country logistic regression models controlling for residence, household wealth, multiple birth status, mother’s education, child’s age and sex, and history of recent fever, odds of anemia were significantly lower for children who used an ITN the night before interview compared to those who did not (OR = 0.81, 95% CI = 0.70–0.94). Decomposition models reveal that the increase in ITN use between baseline and endline surveys accounted for 19% of the observed decrease in moderate-to-severe anemia which is equivalent to a 1.1% reduction in anemia prevalence. The changes in ITN use explained the greatest proportion of the total change in anemia between baseline and endline as compared to other covariates. Results suggest that scale-up of malaria control interventions are likely to have measurable impact on moderate-to-severe anemia and consequently that anemia may be a useful impact measure.