In April, 2012, a 32-year-old white man came to our hospital with fever, lethargy, dyspnoea, and bilateral ankle oedema. He had spent the last 4 months in Togo and had not taken malaria prophylaxis. 15 days previously, he had developed fever and lethargy, and had attended a clinic at which Plasmodium falciparum malaria was diagnosed, with an initial parasitaemia of 1650 para sites per μL. His haemoglobin (Hb) was 128 g/L. He was given 160 mg intramuscular artemether daily for 3 days, followed by a 3-day course of oral dihydroartemisinin/piperaquine, with a good symptomatic response. 2 days later, he had a recurrence of fever, in addition to dyspnoea and ankle swelling. A blood fi lm showed no malaria parasites, but his Hb was 108 g/L. During the next 3 days he deteriorated and was transferred to hospital in Lome. A blood fi lm was again negative for malaria parasites, but the Hb had dropped to 70 g/L. He was prescribed intravenous (IV) ofl oxacin with oral doxycycline and was fl own to London. On arrival at this hospital he was febrile at 38·5°C, tachycardic, clinically anaemic, and jaundiced, had hepato splenomegaly, and was passing dark brown urine. Urinalysis confi rmed moderate haemoglobinuria. His Hb was 41 g/L with a mean corpuscular volume of 83 fL, anisocytosis and mild polychromasia, platelets of 112×109/L, and a white cell count of 4·2×109/L. Reticulocyte count was 5%, haptoglobin <100 mg/L, and direct agglutinin test was negative. Bilirubin was 35 μmol/L. Urea, electrolytes, and liver function tests were within normal limits. Lactate dehydrogenase was raised at 28·17 μkat/L, glucose-6-phosphate dehydrogenase activity was normal at 0·14 nkat/g haemoglobin, haemo globinopathy screen was negative, and iron, folate, and vitamin B12 values were normal. A rapid diagnostic test detecting histidine-rich protein 2 from P falciparum was positive. In-house nested PCR detected P falciparum DNA. However, thick and thin blood fi lms were consis tently negative. A comprehensive infection screen was negative. A fl uorodeoxyglucose positron emission tomography scan was unremarkable. Bone marrow biopsy showed pro nounced erythroid hyperplasia that was consistent with rapid red cell turnover. There was no evidence of any infi ltrative process, haemophagocytosis, or macrophage activation, and perforin expression was normal. No cold antibodies or other atypical auto antibodies or alloanti bodies were detected. A diagnosis of artemisinin induced haemolysis was made, on the basis of having excluded other potential causes of haemolysis. The patient received four blood transfusions over the next 6 days, with evidence of continuing severe intra vascular haemolysis (appendix). On the 7th day after admission, his fever stopped and haemolysis spon taneously resolved. He was discharged home and made a full recovery without further transfusion or treatment and remained well when last followed up in December, 2012. Artemisinin combination therapy has recently become the recommended fi rst-line treatment for both complicated and uncomplicated malaria in endemic areas, and is increasingly used in both Europe and the USA. Artemisinin derivatives are especially valuable for severe malaria because of the rapid rates of parasite clearance, and data from large clinical trials have shown the superiority of IV artesunate over IV quinine. Although the safety profi le of oral and parenteral artemisinins seems excellent in reported clinical trials, severe delayed haem olysis is emerging as a problem in European patients treated with parenteral artemisinins. 10% of patients (seven of 68) treated with IV artesunate in a recent Dutch and Belgian series developed severe haemolysis, with a median lowest Hb of 38 g/L occurring 7–31 days after treatment. Haemolysis persisted for up to 7 weeks. A second European series of 25 patients reported a similar syndrome, with 24% (6/25) developing profound haemo lytic anaemia 14–31 days after IV artesunate. Four additional cases have now been reported from other European centres, and despite extensive investigation in most of these cases, alternative explanations for haemolysis have not been identifi ed. As artemisinin-based therapy becomes the standard of care for severe imported malaria, caution must be taken to monitor for the increasingly recognised and serious com pli cation of severe delayed haemolysis.