This article extends the work of Cohn et al. [1989] on estimating cons
tituent loads to the problem of estimating a percent reduction in load
. Three estimators are considered: the maximum likelihood (MLE), a ''b
ias-corrected'' maximum likelihood (BCMLE), and the minimum variance u
nbiased (MWE). In terms of root-mean square error, both the MVUE and B
CMLE are superior to the MLE, and for the cases considered here there
is no appreciable difference between the MWE and the BCMLE. The BCMLE
is constructed from quantities computed by most regression packages an
d is therefore simpler to compute than the MVUE (which involves approx
imating an infinite series). All three estimators are applied to a cas
e study in which an agricultural tax in the Everglades agricultural ar
ea is tied to an observed percent reduction in phosphorus load. For ty
pical hydrological data, very large sample sizes (of the order of 100
observations each in the baseline period and after) are required to es
timate a percent reduction in load with reasonable precision.