Anaemia is a significant public health issue in many low-income countries, yet little systematic work has examined associated socioeconomic and health factors beyond clinical research. The present study examines the effects of the characteristics of individual children, their mothers and households on anaemia prevalence among Ghanaian children under age five. A generalized linear mixed regression model is applied to data derived from the 2008 MEASURE Demographic and Health Survey in Ghana to estimate fixed and random effects of associated variables on anaemia. Significantly associated factors include child’s age, sex, and fever status in the two weeks preceding the survey, mother’s body mass index, haemoglobin level, health insurance coverage, mother’s education and household wealth status. Factors not significantly associated include whether children received iron supplements, slept under a mosquito bed net and the type of mosquito bed net utilized. Childhood anaemia is related to personal, social and environmental factors. Developing sound policy to address this health problem will require additional research to understand the ways in which these factors are related.