House dust-associated bacterial endotoxins have been shown to be associated with asthma severity, and a similar role has been suggested for fungal (1-->3)-beta-D-glucans. In this study the relation between these agents and peak expiratory flow (PEF) variability was investigated in 148 children 7 to 11 yr of age of whom 50% had self- or parent-reported chronic respiratory symptoms. All children self-monitored twice daily their PEF for a period of 16 wk. Dust samples were collected from mattresses and from living room and bedroom floors, and endotoxin and (1-->3)-beta-D-glucan were measured in dust extracts. The relations with mean daily PEF variability (Ampl%mean) were investigated by linear regression analysis, adjusting for dust mite allergen levels, presence of pets, and type of floor cover. In unadjusted analyses the levels of both endotoxin and (1--> 3)-beta-D-glucan per square meter of living room floor were significantly associated with PEF-variability (but not when expressed per gram of sampled dust), particularly in atopic children with asthma symptoms. Adjusted analyses showed the same association for (1--> 3)-beta-D-glucan but not for endotoxin. Although no associations were found with microbial agent levels in bedroom floor or mattress dust, high levels of (1-->3)-beta-D-glucan in living room floor dust apparently increase PEF variability in asthmatic children.