After seven pediatric cases of chronic fatigue syndrome (CFS) were diagnosed in a farming community in upstate New York, a questionnaire regarding symptoms and potential risk factors of CFS was distributed to all students enrolled in the same school district. Twenty-one students with symptoms of CFS were identified. Two controls per case matched for age and sex were randomly selected from questionnaire respondents. Health status was verified for all subjects by telephone, and diagnosis of CFS was confirmed by a physician. Information was collected on the following factors: symptoms of CFS among other family members; history of allergy/asthma; consumption of raw milk, raw eggs, raw cheese, or raw meat; water supply; exposure to animals; home heating source; proximity to farmland/orchards; tick bite; blood transfusion; camping; and appendicitis. Logistic-regression analyses indicated that the best model (characterized by symptoms among other family members, recent ingestion of raw milk, and history of allergy/asthma) produced significant estimates of relative risk (P less than .05) of 35.9, 44.3, and 23.3, respectively, for the three factors (corrections were made for the effect of the other covariates). These data suggest that a combination of host and environmental factors, including an infectious agent or agents, are involved in the etiology of CFS.