Background Splenic enlargement is a component of the host response to malaria and may also influence the genesis and progression of malarial anaemia. Few cross-sectional and no longitudinal studies have assessed the relationship between splenic volume measured ultrasonographically and haemoglobin concentrations in children with malaria. Methods Fifteen Papua New Guinean children with severe malarial anaemia (SMA; haemoglobin <50 g/L) and ten with moderate malarial anaemia (MMA; 51–99 g/L) were recruited. The SMA patients were given intramuscular artemether followed by oral artemisinin combination therapy (ACT), and were transfused one unit of packed cells 0.3-4.0 days post-admission. The MMA patients were treated with ACT. Splenic enlargement (Hackett’s grade, subcostal distance and ultrasonographically determined volume) and haemoglobin concentrations were measured on days 0, 1, 2, 3, 7, 14, 28, and 42. Results Associations between Hackett’s grade, subcostal distance and splenic volume were modest (rs ≤ 0.62, P <0.001). Baseline splenic volume was not associated with age or haemoglobin (P ≥0.90). Mean splenic volume had fallen by approximately 50 % at day 14 in children with MMA (P ≤0.011 vs days 0, 1 and 2), but there was no change in the SMA group (P ≥0.30). There was no change in haemoglobin in the MMA group during follow-up but a rise in the SMA group to day 7 (P ≤0.05 vs days 0, 1, 2, and 3) which paralleled the packed cell volume transfused. Conclusions Clinical assessment of splenomegaly is imprecise compared with ultrasonography. Serial splenic volumes and haemoglobin concentrations suggest that the spleen does not influence post-treatment haemoglobin, including after transfusion.