The prevalence of allergic disease is low in Eastern Europe for reasons that are poorly understood. Our study aimed to investigate the levels of exposure to indoor allergens and living conditions among Estonian infants in relation to sensitization. Dust samples were collected during four winter months in 1993/94 from the homes of 197 infants participating in a prospective study of sensitization. Information about living conditions was collected through home visit and interviewing the mothers when the children were 6 weeks old. Three dust samples were collected from each home: i.e., from the infant's mattress, bedroom floor, and living-room carpet. The levels of allergens were determined by ELISA with monoclonal antibodies. The highest allergen level in a home was regarded as the peak value. The peak geometric mean values (±SD) of Der p 1 and Der f 1 were 0.3 (0.07–1.4) μg/g dust, of Can 1, 0.86 (0.23–3.12) μg/g dust, and of Fel d 1, 0.1 (0.01–0.9) μg/g dust. In 12 homes (9%), the peak value of house-dust mite (HDM) allergens exceeded 2 μg/g dust, with Der p 1 as the dominating allergen. Multivariate analyses indicated that high levels of HDM allergens were more common in apartments that were on the ground floor or first floor, that were heated with stoves, and/or that had a dampness problem. The mean allergen levels at home were similar in children sensitized to HDM (n = 17. 0.29 v.s 0.3 μg/g dust), dog (n=5, 0.55 vs 1.06 μg/g dust, and cat (n= 18, 0.21 vs 0.09 μg/g dust) and in children who were not sensitized to these allergens. Most of the sensitized children were exposed to relatively low allergen levels at home; i.e., below 1 μg/g dust. This level was exceeded in the homes of 4/17 mite-, 5/18 cat-, and 0/5 dog-sensitized children. The similar levels of the major indoor allergens in Estonia and in Scandinavia indicate that the large differences in atopy prevalence among children and young adults in the two regions are not due to differences in allergen exposure. No allergen threshold level for sensitization was identified.