In a recent article [J. Appl. Phys. 89, 4667 (2001)], Beard applies Taylor dispersion theory for two-dimensional Poiseuille flow between flat plates to analyze mixing processes occurring at the entrance of a microfluidic T sensor. The classical value of the coefficient of the Taylor dispersivity, 1/210, differs from Beard's result, and this Comment points out elements of Beard's derivation which lead to his erroneous value of 33/560. Moreover, the lateral boundedness of the actual device and the long-time asymptotic nature of Taylor dispersion theory call into question the use of the latter model in the context of the T sensor described in Beard's analysis.