Maternal depression has been associated with adverse child growth and development; less is known about its relation to children's diet. In a cross-sectional study embedded at endline of a longitudinal community development intervention, mothers of 629 children (age 23–66 months) in rural Nepal responded to household and children's diet questionnaires and were screened for depression. Child anthropometry and development (Ages and Stages Questionnaire) were assessed. Regression models examined children's diet, growth and development, adjusting for household, child and maternal characteristics. The prevalence of maternal depression was 21%. Maternal depression was associated with 11% lower likelihood that the child consumed one additional food group [Poisson regression, adjusted relative risk (aRR) 0.89, 95% confidence intervals (95% CI 0.81, 0.99), p = 0.024] and 13% lower likelihood that the child consumed one additional animal source food (ASF) [aRR 0.87, (95% CI 0.76, 1.01), p = 0.061] compared with children of nondepressed mothers. However, maternal depression was not associated with either child anthropometry or development: these outcomes were strongly associated with better home child-rearing quality. Stunting also related to child age and intervention group; child development related to mother's education and household wealth. This study suggests a correlation between maternal depression and child dietary diversity. This association could be due to unmeasured confounders, and therefore, further research is warranted. Understanding the relationship of depression to child outcomes—and the role of other potentially compensatory household factors—could help address some of the earliest, modifiable influences in a child's life and contribute to innovative approaches to improve child well-being.