A comparison between two hypothetical flow models of an unconfined aquifer, one saturated and the other variably saturated, indicates that the variably saturated model which explicitly models drainage from the unsaturated zone provides a better conceptual framework for analyzing unconfined aquifer test data and better estimates of the lateral and vertical hydraulic conductivity in fine-grained sands. Explicitly accounting for multiple aquifers, well-bore storage, and the effects of delayed drainage from the unsaturated zone increases confidence in aquifer property estimates by removing some assumptions and allowing for the inclusion of early time data and water-table observations in an aquifer test analysis. The inclusion of the unsaturated zone expands the number of parameters to be estimated, but reasonable estimates of lateral and vertical hydraulic conductivity and specific storage of the unconfined aquifer can be obtained. For the cases examined, only the van Genuchten parameter a needed to be determined by the test, because the parameters n and 9r had a minimal effect on the estimates of hydraulic conductivities, and literature values could be used for these parameters. Estimates of lateral and vertical hydraulic conductivity using MODFLOW were not as good as the VS2DT based estimates and differed from the known values by as much as 30 percent.