Spatial and statistical variability in soil characteristics must be addressed in using soil testing to guide precision nutrient management. This paper provides a case study comparing strategies using technology currently available and economically viable for farmers or their advisors to use for this purpose. The first strategy is to divide a large area into smaller subunits for sampling, then to map results by assigning the soil test value for each sample to the entire subunit, resulting in a mosaic of soil test values across the entire sampling area. An alternative approach involves collecting soil samples from known locations using global positioning system (GPS) technology, then mapping the spatial distribution of soil test results. The final strategy is to use the cumulative distribution function (CDF) to find the percentage of samples with soil test values at or below certain levels irrespective of their location. Based on 72 soil samples from a highly variable 40 ha research site, we found that inaccuracy of GPS may limit its application. Maps communicate soil test results readily, but may be difficult to apply in fertilizer management. The CDF approach provides useful information, but interpreting and applying the information may be difficult. Any of these methods of assessing soil test variability will require analysis of far more samples than composite sampling, and the value of the added information must justify increased analytical costs.