We represent results of a laboratory investigation of the attenuation of humidity fluctuations in tubing. The purpose is to obtain simple formulations for use in turbulent flux measurement applications, where the measurement of a trace species may be remote from the sampling point so that the sample must be transported through a tube to the sensor. The laboratory data are compared to the attenuation predicted from Taylor's (1953 and 1954) formulations for the virtual longitudinal eddy diffusivity of a constituent introduced into a pipe, for both laminar and turbulent flow, by plotting the normalized half-power frequency against Reynolds number. Results were first obtained for fully developed equilibrium flow (both laminar and turbulent) through a straight tube. The following modifications were then studied: an intake section, a 90¿ elbow bend, and curved tubes (bending through 90¿) with tube-radius to radius-of-curvature ratios of a/R=0.034 and 0.081. The effect, if any, of these modifications is to increase the half-power frequency over that of the equivalent length of straight tubing. Applying this analysis to a situation where a sample is collected on a tower and transported to an instrument at the base of the tower, we find that as long as the flow in the tube is turbulent, the attenuation in the tubing should not be a significant factor for introducing errors in flux measurements in the unstably stratified atmospheric surface layer. ¿ American Geophysical Union 1991