Objective: To investigate haematological and biochemical iron indices in relation to malaria, gravida, and dietary iron status in rural pregnant Malawian women. Design: In this self-selected sample, haemoglobin, haematocrit, red cell indices, serum ferritin, serum iron, serum transferrin, and serum transferrin receptor (TfR) were measured. Infection was assessed by a malaria slide, serum C-reactive protein, and white blood cell count. Dietary iron variables were measured by three 24-h interactive recalls. Setting and subjects: 152 rural pregnant women recruited at 24 weeks gestation while attending a rural antenatal clinic in Southern Malawi; 36% were primagravid; 43% were gravida 2 ± 4; 26% were gravida b 5. Results: Of the women, 69% (n 105) were anaemic (haemoglobin`110 gal); 37% (n 39) had anaemia and malarial parasitaemia on the test day; 17% (n 26) with malaria were also classi®ed with iron de®ciency (ID) anaemia (based on serum ferritin 50 mgal and Hb`110 gal) while an additional seven with malaria were classi®ed with ID without anaemia. In malarial-free subjects, 32% were classi®ed with IDA (serum ferritiǹ 12 mgal and Hb`110 gal) and 17% with ID (serum ferritin`12 mgal; Hb ! 110 gal). Serum TfR concentrations were elevated in anaemic women (P`0.01). In non-malarial parasitaemic subjects, serum TfR correlated negatively with haemoglobin (r 7 0.313; P`0.001) but not serum ferritin. Of the women, 49% were at risk for inadequate iron intakes. Most dietary iron was non-haem; plant foods provided 89%;¯esh foods (mainly ®sh) only 9%. Malarial parasitaemia and intakes of available iron impacted signi®cantly on iron status. Conclusion: Anaemia prevalence from all causes was high (that is, 69%); three factors were implicated: malaria, and de®ciencies of iron and possibly folate, induced partly by an inadequate dietary supply andaor secondary to malarial parasitaemia.