1. Transnasal pressure (delta p) in Pa and gas flow in cm3 s-1 were measured in men breathing gases of different density, assuming flow through each nostril (f) to be half the measured total flow. Measurements were also made in children with allergic rhinitis before and after nasal antigen challenge. 2. Flow always showed evidence of turbulence when transnasal pressure exceeded 40-80 Pa breathing air and Reynolds number exceeded 2400 (1800 in the presence of nasal obstruction). 3. Changes in effective mean nasal airway diameter (D) after nasal challenge can be determined at points of similar pressure from the relation log (D2/D1) = 0.368 log (f2/f1) when flow is turbulent. 4. Absolute estimates of airway diameter in cm can be obtained from the relation 4.75 log D = 1.75 log f--log delta p+log L--4.756 breathing air when flow is turbulent (since the relation between pressure and flow approximates to that found in a long cylinder) and these estimates are only marginally affected by differing assumptions about nasal airway length (L). 5. Because pressure and flow are non-linearly related, and changing nasal dimensions have a profound effect on the flow at which turbulence occurs, it is suggested that measures of nasal conductance at delta p greater than or equal to 0.1 kPa are preferable to the more conventional measures of nasal resistance at a specified flow rate.