Objective: Children from low-income countries consuming predominantly plant-based diets but little animal products are considered to be at risk of Fe deficiency. The present study determined the Fe status of children from resource-limited rural households. Design: A cross-sectional study. Setting: Twenty six kebeles (the smallest administrative unit) from six zones of the Amhara region, Ethiopia. Subjects: Children aged 54-60 months (n 628). Results: Grain, roots or tubers were the main dietary components consumed by 100 % of the study participants, followed by pulses, legumes or nuts (66·6 %). Consumption of fruit and vegetables (19·3 %) and meat, poultry and fish (2·2 %) was low. Children had a mean dietary diversity score of 2·1 (SD 0·8). Most children (74·8 %, n 470) were in the lowest dietary diversity group (1-2 food groups). Rate of any morbidity in the preceding 14 d was 22·9 % (n 114). Infection or inflammation (α 1 -acid glycoprotein >1·2 g/l) was present in 30·2 % (n 184) of children. Children had a high rate of stunting (43·2 %). Of the total sample, 13·6 % (n 82) of children were anaemic, 9·1 % (n 57) were Fe deficient and 5·3 % (n 32) had Fe-deficiency anaemia. Fe-deficiency erythropoiesis was present in 14·2 % (n 60) of children. Conclusions: Despite consuming a predominantly plant-based diet and little animal-source foods, there was a low prevalence of Fe-deficiency anaemia. This illustrates that dietary patterns can be inharmonious with Fe biochemical status; thus, Fe-related interventions require biochemical screening.