The occurrence of dog hypersensitivity in 203 unselected asthmatic children was investigated by means of skin prick test, provocation test and RAST. The history of past and present exposure to dogs, symptoms in contact with dogs, and the effects of dog avoidance were recorded. The amount of dog dander antigens was measured from dust samples from 67 homes using counter-immunoelectrophoresis. A history of past or present dog hypersensitivity was obtained for 120 subjects (59%). The skin prick test was found to be a sensitive method for diagnosing dog hypersensitivity, when a standardized allergen extract was used. The results of the skin prick test correlated well with the provocation test (r = 0.832) and the RAST (r = 0.777). A positive (greater than or equal to ++) skin reaction was observed in 56% of the subjects, a positive (class greater than or equal to 1) RAST in 69%, and a positive provocation test with the same extract in 41% of the subjects. The occurrence of dog hypersensitivity was not significantly associated with past or present exposure to dog at home. However, the children who were exposed to a dog at home during early childhood developed dog hypersensitivity symptoms significantly earlier and had dog hypersensitivity slightly more often than the subjects with a later exposure or no exposure to dog. Significant amounts of dog dander antigens were found even in dust samples from homes where dogs had never been kept. The effect of dog avoidance on dog dander-specific IgE levels was studied from sera obtained from 24 subjects. Steadily high and even rising levels were observed in cases when a strict avoidance of dogs was reported by the patient. Dog saliva was found to be a more potent and more species-specific source of dog allergens, in comparison with a commercial dander extract, according to skin prick test and RAST inhibition studies.(ABSTRACT TRUNCATED AT 250 WORDS)