The insertion of a suction catheter or a bronchoscope down an endotracheal tube increases the resistance to gas flow down the tube. The extent to which this occurs depends on the relative diameters of the endotracheal tube and the coaxially introduced catheter. This study utilises a laboratory model to quantify this effect, using a steady flow down an annulus between two tubes whose long axes lie co-axially. Two diameters of an endotracheal tube were modelled to represent flow down adult and neonatal endotracheal tubes; these were of internal diameter (d(o)) 6.3 mm and 3.2 mm, and of length (L) 555 mm. A steady flow of air was generated to pass through the model 'endotracheal' tube. Flowrates were calculated to give Re of approximately 5000 for the larger endotracheal tube, and of approximately 1300 for the smaller. These values correspond to clinically appropriate flowrates in adult and neonatal patients, respectively. The pressure drop deltaPo down the endotracheal tube was measured initially without any obstruction, using a calibrated pressure transducer. Catheters of diameter (d(i)) 0.8 mm, 1.6 mm, and 3.2 mm were introduced into the larger diameter endotracheal tube, while catheters of 0.8 mm and 1.6 mm were introduced into the smaller one, and flow was restored to its original value. The pressure drops deltaP down the endotracheal tubes were measured with the catheters introduced a length 'x' into the tube, to x = L/2 and to x = L. Results are compared with a theoretical calculation on the basis of laminar flow for concentric tubes. If a sampling tube or suction catheter is used down the length of an infant's endotracheal tube, the results show that for most values of do/di, there is a significant rise in deltaP/deltaPo. Where a flexible bronchoscope is used down an endotracheal tube or a telescope down a rigid bronchoscope, the value of deltaP/deltaP(o) may also increase unacceptably where d(o)/d(i) is low. The results show that for equal d(o)/d(i), and equal values of x, deltaP/deltaPo are lower for higher values of Re than for lower; and that for lower values of Re there is a more rapid increase in deltaP/deltaPo as x increases, than for higher Re, especially at low values of d(o)/d(i). This result quantifiably confirms clinical experience; that care must be taken in introducing a catheter down a neonatal endotracheal tube. Deviation of these results from the theoretical calculation is less for the smaller Reynolds numbers and smaller values of d(o)/d(i), because under these conditions the flow is more likely to be laminar, with a greater degree of concentricity.