The relationship among inhaled allergen exposure, sensitization, and asthma severity is unknown. To investigate the relationship among personal allergen exposure, reservoir dust allergen concentrations, and physiological measures of asthma severity; to examine the numbers of particles inspired that react with autologous IgE and IgG4. A total of 117 patients with asthma wore 5 nasal air samplers (NASs) at home: 1 each for exposure to mite, cat and dog allergens, NAS-IgE, and NAS-IgG4. NASs were processed by HALOgen assay for allergen measurement and incubated with autologous serum for detection of NAS-IgE and NAS-IgG4. Reservoir allergen concentrations were measured by ELISA. Subjects' asthma severity was ascertained by measurement of lung function, exhaled nitric oxide, and nonspecific bronchial reactivity to histamine. Nasal air sampler counts correlated with reservoir concentrations for cat (r=0.31; P=.001) and dog (r=0.20; P=.03) but not mite allergen (r=0.001; P=1.0). There was no significant relationship between sensitization with exposure measured by NAS to any allergen and PD20FEV1 (F[3,60]=1.60; P=.20); however, sensitization with exposure in dust reservoirs had significant effects on PD20FEV1 for any allergen (F[3,59]=3.12; P=.03), cat (F[3,59]=3.77; P=.01), and mite (F[3,59]=2.78; P=.05), but not dog (F[3,59]=1.06; P=.37). We repeated the analysis with separate variables for sensitization and exposure, controlling for the confounders; sensitization but not exposure conferred lower PD20FEV1 values. However, increasing cat allergen exposure was associated with improving bronchial reactivity in not cat-sensitized patients. NAS-IgE and NAS-IgG4 counts bore no relationship to any measure of asthma severity. Nasal air samplers confer no advantage over reservoir dust analysis for studies of asthma severity. In common with other measures of exposure, single nasal air samples do not provide a useful measure of home allergen exposure for the individual patient with allergic asthma.