Similarities between the geographical and climatic prevalences of kwashiorkor and of exposure to dietary aflatoxins, and between the biochemical, metabolic and immunological derangements in kwashiorkor and those in animals exposed to aflatoxins, prompted investigation of the associations between kwashiorkor and aflatoxins. Studies in Africa in the 1980s indicated a role for these toxins in the pathogenesis of the disease. Paediatric cases of kwashiorkor are less prone to severe Plasmodium falciparum malaria than normal children. In mice infected with P. berghei, aflatoxin exposure inhibits parasite growth and ameliorates morbidity. Aflatoxins occur in < or = 40% of samples of breast milk from tropical Africa, usually as low concentrations of the relatively non-toxic derivatives of aflatoxin B1 (AFB1) but sometimes as high concentrations of the very toxic AFB1. This could explain kwashiorkor in breast-fed babies. Aflatoxin exposure occurs in > or = 30% of pregnancies in tropical Africa and the toxins are often in cord blood, sometimes at extremely high concentrations. Aflatoxins are now incriminated in neonatal jaundice and there is circumstantial evidence that they cause perinatal death and reduced birthweight. Aflatoxin-induced immunosuppresion may explain the aggressive behaviour of HIV infection in Africa. There are similarities between observations on HIV cases in Africa and those on heroin addicts in Europe, where 'street' heroin is frequently contaminated with aflatoxin. Aflatoxins were found in 20% of random urine samples from heroin addicts in the U.K. and the Netherlands. Aflatoxins have also been incriminated in episodes of food poisoning which have been associated with serious morbidity and mortality, particularly among young children.