Factors involved in the development of inflammation and asthma in nonatopic subjects have remained largely obscure, although there is some evidence to suggest that certain infections may play a role. We investigated the association between serological evidence of Chlamydia pneumoniae infection and asthma in adults, and the possible modifying effect of the patients' atopic status on this association. Four hundred and thirty consecutive patients who attended the hospital between 1992 and 1993 with symptoms suggestive of asthma, rhinitis or allergy were enrolled. Diagnostic procedures including lung function measurements and skin-prick tests were performed in all patients. The patients with established asthma (n = 332) were divided into those with recent asthma (n = 224, onset 1985 onward) and longstanding asthma (n = 108, onset before 1985). The controls (n = 98) comprised all subjects who did not meet the criteria of asthma. Serum immunoglobulin (Ig)G, IgA and IgM antibodies to C. pneumoniae were measured by the microimmunofluorescence test. In women, the prevalences of elevated IgG (a titre of >/= 128) and IgA (>/= 32) antibody levels and the age-adjusted geometric mean titres (GMT) of IgG and IgA antibodies were invariably highest among subjects with nonatopic longstanding asthma. Elevated IgG titres in women occurred in 11% of controls, in 28% of nonatopic recent onset asthmatics, and in 43% of asthmatics with nonatopic longstanding disease; for men the respective figures were 33, 50 and 64%. Logistic regression analysis controlling for age, sex and smoking showed that asthma was significantly associated with elevated IgG antibody levels to C. pneumoniae (odds ratio 3.3, 1.6-6.8 for longstanding asthma, 2.3, 1. 2-4.4 for recent asthma, and among women only 4.2, 1.6-10.9 for longstanding asthma, and 3.0, 1.3-7.2 for recent asthma). When the atopics and nonatopics were analysed separately, an even stronger relationship in the nonatopics was obtained for longstanding asthma (6.0,2.1-17.1). In contrast, the relationship between atopic asthma, either recent or longstanding, and elevated IgG titres was not significant, indicating that asthma per se does not predispose to C. pneumoniae infection. Asthma was significantly associated with elevated IgG antibody levels to C. pneumoniae, and this association was strongest for nonatopic longstanding asthma.