In April, 2006, a 32-year-old man was referred to us by his local primary health-care clinic, with a cough that had been present for 2 weeks, accompanied by mild haemoptysis and diff use pleuritic chest pain. He was from Mozambique, but had been working in Johannesburg—initially as a labourer on building sites, and then as a gardener. He did not smoke. He had been diagnosed as HIV positive in 2005. On examination, the patient was not underweight or short of breath. He had cervical, axillary, and inguinal lymphadenopathy. We heard coarse crackles over both lungs, and palpated the liver edge 8 cm below the costal margin. We found no splenomegaly or ascites. Chest radiography (fi gure) showed many round opacities. Ultrasonography of the abdomen showed many hypoechogenic lesions in the liver. Blood tests showed that the patient was mildly anaemic, with a normal white-cell count. However, the CD4 cell count was only 27 per μL. Serum concentrations of alkaline phos phatase and γ-gluta myltransferase were raised, at 335 U/L and 228 U/L respectively. Serological testing gave a positive result for hepatitis C. CT of the chest and abdomen showed many cystic lesions in the lungs and in the liver; an indirect haemagglutination assay was positive for Echinococcus spp, at a titre of 1:512; a titre of 1:20 is regarded as diagnostic. We prescribed albendazole, at a dose of 400 mg twice daily, and prednisolone. The patient was discharged after 2 weeks, and completed the course of albendazole, which consisted of three 4-week periods of medication, separated by 1-week intervals intended to reduce the risk of liver damage. In August 2006, repeat CT of the chest and abdomen revealed that the cysts had decreased in size and number. We did not start treatment with highly active antiretroviral drugs while the patient was taking albendazole, because of the possible drug interaction, the risk of iatrogenic liver damage, and the risk of immune reconstitution syndrome. The patient was counselled about the risks and management of HIV; we intended to prescribe antiretroviral drugs after he completed the course of albendazole. However, the patient did not attend follow-up appointments, and attempts to trace him were unsuccessful. Echinococcus is a tapeworm carried by dogs. As a gardener, our patient was at increased risk of unwittingly swallowing the tapeworm’s eggs, by which means people become infected; the peak age of infection is 30–40 years. After being swallowed, the eggs release embryos, which travel in the bloodstream before lodging in organs such as the liver and lungs, where the embryos develop into cysts containing larvae. Echinococcus cysts are also known as hydatid cysts. Most cysts are asymptomatic; indeed, they are often found serendipitously, on radiological testing. Patients with AIDS are more likely to have echinococcal disease that develops rapidly and manifests early. However, our patient presented with mild symptoms, despite disseminated cystic disease and infection with hepatitis C. We speculate that perhaps the mildness of the symptoms was caused by the impairment of his immune response. Although our patient accepted treatment for tapeworm infestation and HIV counselling, he did not attend appointments for HIV treatment. He may have feared unemployment; his health beliefs may have caused him to doubt the diagnosis or treatment of AIDS; the grief or stigma of the diagnosis may have prevented him from accepting it.