Studies on estimation of actual evapotranspiration on disturbed lands are scarce and yet such data are essential in hydrologic modeling. Our study compared the variability of estimates of actual evapotranspiration (AET) from a reclaimed site in northern Alberta using the simplified water budget (WB) and soil water reduction (SWR) methods. The AET estimates from the simplified water budget equation (AET1) required field soil water content, precipitation and runoff. The AET estimates from the soil water reduction method (AET2) required daily potential evapotranspiration (PET), field capacity (FC) water content, minimum field-measured water content, and field water content. Soil water was measured using neutron moisture gauge every 2 wk during the growing season for a 2-yr period. The average AET1 and AET2 estimates for 2001 were 1.9 and 1.4 mm per day, respectively, where as those for 2002 were 2.1 and 1.2 mm per day, respectively. The paired t-tests to compare AET1 against AET2 indicated significance differences (P ≤ 0.05) in 5 out of 11 measurement dates, especially during high rainfall periods. Overall AET2 estimates were more variable than AET1 estimates. Therefore, the larger variability of AET2 estimates imply less reliability of spatially averaged AET2 estimates for use in regional and global circulation models.