On the basis of recent work, it would appear that the transport coefficients descriptive of certain idealized cases of coupled flow of immiscible fluids in porous media, can be determined in principle by calculations employing well-defined experimental data. Other considerations show, however, that even small errors inherent in laboratory observations sometimes will have an enormously large effect on the accuracy of the calculated values. On the other hand, it often will be the case that values for the transport coefficients are not needed individually, as long as those particular functions which appear lumped together in the equations of motion, can be calculated from the same data but with less error. In any case, it will be clear that error problems will be, to some extent, mitigated when very accurate instrumentation is available to control and measure the fluxes and driving forces that give rise to the transport processes under study. Thus, the aim of this paper is to present an error analysis that will facilitate laboratory design in preparation for experimental work, and will also facilitate the interpretation of the data that eventually are to be obtained. Another aim is to underscore the risks taken by those who fail to take coupling effects into account just because the potential importance of them is not appreciated or not clearly revealed by existing data.