Summaryobjective To determine the relative risks of socio-demographic, dietary, and environmental factors for endomyocardial fibrosis (EMF) in Uganda.method Unmatched case control study in Mulago Hospital, Kampala. Cases (n ϭ 61) were sequential patients hospitalized with an echocardiographic diagnosis of EMF from June 1995 to March 1996. Controls (n ϭ 120) were concurrent patients with other forms of heart disease (heart controls, n ϭ 59) and subjects admitted for trauma or elective surgery (hospital controls, n ϭ 61). All consenting subjects answered a structured questionnaire administered by trained interviewers. Complete blood counts, malaria films and stool examination for ova and parasites were performed. Questionnaires elicited information on home address, economic circumstances, variables concerned with environmental exposures and usual diet before becoming ill.results After adjustment for age and sex, cases were significantly more likely than controls to have Rwanda/Burundi ethnic origins (P ϭ 0.008). Compared with controls, cases had a lower level of education (P Ͻ 0.001 for heart controls and P ϭ 0.07 for hospital controls), were more likely to be peasants (P Ͻ 0.001), and to come from Luwero or Mukono Districts (P ϭ 0.003). After further adjustment for peasant occupation, cases were more likely than controls to walk barefoot (P ϭ 0.015), consume cassava as their staple food (P Ͻ 0.001) and to lack fish or meat in dietary sauces (P ϭ 0.02). Cases were more likely to exhibit absolute eosinophilia (P ϭ 0.006). The effect of cassava diet was more marked in the younger age group, while the effect of eosinophilia was greater in adults. Socio-economic disadvantage is a risk for EMF. Absolute eosinophilia is a putative cause of EMF, a finding not explained by parasitism.conclusion Data indicate that relative poverty and environmental factors triggering eosinophilia appear to act in a geographically restricted region of Uganda in the aetiology of EMF. keywords endomyocardial fibrosis, cassava, eosinophilia correspondence Prof. J. L. Ziegler,