Variable methods of dust collection may lead to uncertainty in the measurement of biomarkers. The purpose of this study was to examine the effect of two different dust collection devices on dust weight, Der p 1, Fel d 1, and endotoxin levels. We compared: (1) a nylon mesh sock inserted between the furniture attachment and the vacuum hose (the reference method) and (2) the ALK device. Duplicate dust samples were collected for 2 min from 2 m(2) of 37 living room floors and from each longitudinal half of 37 mattresses. Measurement of Der p 1 and Fel d 1 were by double monoclonal antibody enzyme-linked immunosorbent assay (ELISA) and endotoxin by a Limulus Amobocyte Lysate assay. Geometric mean ratios (95% confidence intervals) were calculated to show the differences between sampling devices for each measurement. Compared with the ALK device, the reference method collected significantly more dust from floors (sevenfold) and mattresses (threefold) and more total Der p 1, Fel d 1, and endotoxin in both sites. Floor, but not mattress, Der p 1 concentrations were also significantly higher (threefold) using our reference method. We recommend that, in order to minimize sampling device bias, allergen and endotoxin are expressed as a concentration, and that the bed is considered the major source of allergen exposure. Practical Implications Dust sampling equipment can influence the dust yield. In order to have confidence in comparisons of allergen and endotoxin reservoir levels between centers, standardization in the use of sampling equipment is important.