Household dust samples from the homes of 106 allergy clinic patients in Baltimore were analyzed for specific allergen content. Dust mite antigen content was determined by enzyme-linked immunosorbent assays (ELISA) specific for the major allergens of Dermatophagoides pteronyssinus, D. farinae, and D. microceras. Cat and dog antigen content were determined by ELISA using antisera for Fel d 1 (formerly cat allergen 1) and dog allergens 3 and 13, respectively. Mold content was assessed by culture with microscopic identification. Dust mite antigen was detected in 99% of homes (D. farinae, 95%; D. pteronyssinus, 88%; D. microceras, 31%), with total antigen content ranging from 50 ng/g dust (the lower limit of detection) to 30,170 ng/g (median, 1,123 ng/g). Animal allergens were found in 100% of samples (cat: range, 2 to 130,000 ng Fel d 1/g; median, 90 ng/g; dog: range, 112.5 to 585,000 IU/g; median, 2,719.5 IU/g). Although there were highly significant differences in antigen content (p less than 0.001) between homes with and without a particular pet in residence, many homes without pets contained pet allergens at high concentrations. Molds were also detected in 100% of homes (range, 4 to 761 colonies/30 mg dust; median, 72 colonies/30 mg). No correlation was demonstrated between antigen content and skin test results, a history of asthma, symptoms on allergen exposure, or the age of the home (except for molds) for any of the allergens detected. We conclude that dust mite allergens, cat and dog allergens, and molds are virtually ubiquitous in Baltimore homes and that our ability to predict the presence and relative quantities of these allergens on clinical grounds is very limited.