Knowledge of soil water fluxes is critical for evaluating efficiency and environmental effects of soil and crop management. Indirect methods commonly used to estimate soil water fluxes estimates are currently based on (a) soil water balance, (b) soil water potential measurements with the Darcy-Buckingham law application, (c) simulations of soil water transport with the Richards equation, and (d) capillary lysimeters. The methods differ in their practicality and certainty of results. The Gee Fluxmeter is a direct method that has been developed but not widely tested. The objective of this study was to compare water flux estimates obtained from these methods and to evaluate the Gee Fluxmeter against these results. The study was conducted within a highly controlled lysimeter (21.3 m x 13.7 m x 3.0) with a sloping surface soil and instrumented to provide for real-time and near continuous measures of the water table, soil water content, soil water potential, meteorological parameters, and solute flux. Important variables that could not be directly measured were runoff and evapotranspiration. Each method was associated with its own sources of error and instrumentation malfunctions resulted in some missing data that had to be estimated from other sources. The water balance method (a) depends on flux estimates from method (b) so that the calculated fluxes from these two methods are not independent (when hydraulic conductivity is not known). Short-term estimates (rainfall event based) of fluxes from the indirect estimates were different from each other. The Gee Fluxmeter worked well but the capacity was exceeded for rainfall events with large amounts of recharge. The correspondence between the direct and indirect methods improved as the time scale increased.