The value of soil test information in planning application levels of f
ertilizer is determined by using agricultural field plot data to estim
ate the corn production function, to construct plausible prior distrib
utions of nitrate levels and to determine the sampling distribution of
soil test results. These quantities are used to determine the posteri
or distribution of soil nitrate concentrations given the results of a
soil test Optimal decisions concerning application levels of fertilize
r are made with respect to this posterior distribution. Average reduct
ions in fertilizer application rates range from 15% to 41%, depending
on the form of prior information that is available. These reductions a
re achieved by increasing the variability of application rates over ti
me. Disregarding the uncertainty that remains after soil testing signi
ficantly overstates the expected benefits of soil testing.